Indranil Bhattacharya

Research | Algorithm Design and Analysis | Machine Learning | Deep Learning | Information Retrieval | Forecasting | Gen-AI - Working with the Applied AI Science team. Our vision is for knowledge operations work across Worldwide Amazon Stores (WWAS) to be automated, enabling teams to focus on the most creative and critical decision-making work, while raising the quality bar for outcomes. Our mission is to build reusable solutions that delight organizations in WWAS who leverage AI with the goal of making knowledge operations work more efficient. - Worked on Shipments Per Route (SPR) forecasting problem using Physics-assisted Machine Learning for Last Mile Planning Science team - Worked on task decomposition and planning using LLM and IR methods (RAG) and search methods (MCTS, BFS, DFS) - Worked on building AutoML solutions @Amazon scale thereby enabling non-tech users to self-serve build, train, deploy and maintain ML models (for business policy driven use cases) in production that operate on products hosted in the Amazon catalog. - Worked in the area of Robust Learning at catalog scale. - Worked on Batch mode Target-Aware Active Learning at catalog scale. - Worked on devising Machine Learning solutions for several problems related to Focused Web Crawlers. - Developed expertise in content/structure/position based classification/characterization of web documents using both classical machine learning and modern day deep learning techniques with strong emphasis on CV, NLP. - Prior experience of working in Causal analysis of time series using Granger Causal models.