• Built NLP solutions using Generative AI models including Gemma and Llama 3.• Built a LLM based chatbot to play the 20 Questions game. The solution ranked among the top 30 or top 3% of the teams globally.• Developed Statistical and Machine Learning models to help US based clients in forecasting and investing.• Implemented different Statistical techniques for detection of multivariate outliers to help in investing. • Worked on a web application to connect farmers and consumers… Show more • Built NLP solutions using Generative AI models including Gemma and Llama 3.• Built a LLM based chatbot to play the 20 Questions game. The solution ranked among the top 30 or top 3% of the teams globally.• Developed Statistical and Machine Learning models to help US based clients in forecasting and investing.• Implemented different Statistical techniques for detection of multivariate outliers to help in investing. • Worked on a web application to connect farmers and consumers directly. • Used SoTA Machine Learning techniques in the areas of Marketing, Finance, NLP and Computer Vision.• Presented and Published papers at prestigious, peer reviewed international conferences on Artificial Intelligence.• Won medals in Kaggle competitions, consistently ranking among the top Data Science competitors Globally.• Evaluated essays using language proficiency models including an ensemble of LGBM, DeBERTa family of models and HuggingFace transformers. This helped improve automated feedback provided to students.• Built portfolios of stocks for the Japanese Stock Exchange Group (JPX). Close to 2000 stocks were ranked based on their predicted returns. Models used include LGBMs, Decision Tree Regressors among others.• Predicted e-commerce clicks, cart additions, and orders by building a Multi-Objective Recommender System for Otto – Germany’s largest online shop. This helped the retailer in providing better recommendations to the customers while increasing sales for online retailers. Models like Biased random walk and XGBoost were used. Show less

• PhD Research on Semantic Search - Web Search based on meaning, rather than words, using real world text and hyperlink data. • Completed PhD coursework from IIT Bombay with a CGPA of 7.74/10.• Research on 'chatbots' or conversational agents to automate Question Answering.• Building Ensembles of meta-algorithms and advanced statistical models. Implementation of Machine• Learning algorithms like Random Forests, Gradient Boosting Methods, Deep Neural Networks including BERT and… Show more • PhD Research on Semantic Search - Web Search based on meaning, rather than words, using real world text and hyperlink data. • Completed PhD coursework from IIT Bombay with a CGPA of 7.74/10.• Research on 'chatbots' or conversational agents to automate Question Answering.• Building Ensembles of meta-algorithms and advanced statistical models. Implementation of Machine• Learning algorithms like Random Forests, Gradient Boosting Methods, Deep Neural Networks including BERT and its variants, CNN, RNN and LSTM.• Writing and executing code on the PARAM YUVA II HPC and on Clouds with GPUs/TPUs.• Automation of Time Series modeling using meta-algorithms from cutting edge research papers.• Research on implementation of Blockchain to different industries.• Correctly predicted the end of 'Big Data', way back in 2015.• Mentoring of junior Professors in the aforementioned areas.• Designed a mobile based e-commerce application to enable a Web Based Re-intermediation of the• Commodity Distribution Channel, for the Small and Marginal Farmer.• Principles of Agile, DevOps, Scrum and Software Requirements Specifications were applied.• Data Science Advisor to Companies founded at Indian Institute of Technology (IIT Bombay) and outside. Show less

• Worked with Large Amounts of Data from a Variety of Database Systems.• Built Statistical Modeling Tools with Various Programming Languages.• Applied Advanced Statistical modeling, Machine Learning and Optimization Techniques to Solve Problems.• Involved in Analysis of Financial Statements, Actuarial Modeling and Derivatives Pricing.• Generated New Product Development Ideas.• Automated Processes, Mentored and Managed People.• Built the SAS Scorecard System – an… Show more • Worked with Large Amounts of Data from a Variety of Database Systems.• Built Statistical Modeling Tools with Various Programming Languages.• Applied Advanced Statistical modeling, Machine Learning and Optimization Techniques to Solve Problems.• Involved in Analysis of Financial Statements, Actuarial Modeling and Derivatives Pricing.• Generated New Product Development Ideas.• Automated Processes, Mentored and Managed People.• Built the SAS Scorecard System – an automated modeling framework built to be both Repeatable and Scalable to generate ‘n’ number of models within the same amount of time it takes to build just one.• Statistical techniques applied include Clustering, Market Basket Analysis/Association Rules, Segmentation,Decision Trees, Time-Series methods, Regressions for Credit Scoring - including Scaling, calibration and validation of Risk/Application Scorecards, Stress testing, and Scenario Analyses.• Used Marketing Mix Modeling to forecast and plan the impact of the choice of marketing channels. • Provided a new Strategic direction to the Fraud Analytics/ Anti-Money Laundering (AML) function at GE, while mentoring employees from USA, Latin America, China, Europe and Australia. • Introduced new software tools; used, deployed and encouraged the use of Open-source software to build additional data analytics capabilities. • Used new tools in parallel with legacy ones to ensure stable platform transformation. Show less