I am currently a Staff Software Engineer in LinkedIn's AI team, primarily focusing on challenging cross-product problems. I lead several efforts in AI that can be applied to different product applications. Currently, my primary focus revolves around extreme-scale linear programs and enhancing fairness in AI.
Through the years at LinkedIn, I have worked on a variety of problems and on various product applications. My focus has ranged from developing prediction models for complex recommender systems powering News Feed Ranking and People You May Know (PYMK) to extreme large-scale optimization problems trying to solve complex matching and allocation problems. I've been the chief architect and designer for the AutoML library used internally by various teams such as Feed, Notifications, Ads and PYMK. I've also worked towards developing accurate causal estimates in the presense of network interference.
Before joining LinkedIn, I finished my PhD in the Department of Statistics at Stanford University advised by Prof. Art Owen. My dissertation was on Quasi-Monte Carlo Methods in Non-Cubical Spaces. Prior to the doctoral program, I earned my B.Stat (Hons) and M.Stat (Specializing in Mathematical Statistics and Probability) from Indian Statistical Institute, Kolkata.
- Large-Scale Optmization
- Causal Inference on Networks
- Fairness in AI
- Monte Carlo Methods
- Statistical Machine Learning
- Bayesian Optimization and AutoML
- Discrepancy Theory
- Numerical Techniques
Outside of work, I enjoy fast cars, painting, playing the sitar, photography, and traveling the world with my wife Devleena Samanta.