Kartik Sreenivasan

WORK HISTORY

EDUCATION

SKILLS

ABOUT KARTIK SREENIVASAN

I am interested by complex problems in Artificial Intelligence, specifically those involving Optimization and Machine Learning. Picking up concepts quickly, rapid prototyping and being comfortable with a breadth of technologies are my strengths.I am currently wrapping up my PhD at UW-Madison in Computer Science focussing on Large-Scale Machine Learning. I have worked on problems related to Adversarial Robustness of Federated Learning, the Lottery Ticket Hypothesis of Deep Neural Networks and Contrastive Learning. For the last few years, I have been focussing on better understanding Large Language Models.At Adobe, I worked primarily on using Machine Learning to improve the efficacy of Search Engine Marketing. My work involved building mathematical models to predict the clicks, and cost of a given ad on a search engine accurately. I have also worked on the problem of choosing optimal bids to maximize the revenue of a given portfolio while constrained to a budget. All these problems are sensitive to Big (potentially sparse) data as well as the performance complications that arise due to scale.Graduated with B.Tech in Information Technology as Gold Medalist at NITK - formerly KREC -(ranked 8th among India\'s engineering schools).