I’m a Software Engineer at Google, working on large-scale recommendation systems. Previously, I finished my PhD in Electrical Engineering & Computer Science (core focus: Machine Learning) from Massachusetts Institute of Technology.

I have extensive experience in machine learning for developing models and algorithms for 

• Neural Training of Additive Models and Tree Ensembles

  • Generalized additive models under sparsity and structural constraints for interpretability

  • Flexible and efficient tree ensemble with application tailored loss functions, multi-task learning etc.

  • End-to-end feature selection for learning skinny tree(s).

• Sparse Mixture of Experts and Routing Approaches for more efficient training and inference of large vision and language models. 

• Pruning large vision and language models for efficient inference.

• Time series forecasting with graph neural networks 

I have a proven track record of publications that highlight creative thought in developing novel research ideas, design of experiments and academic writing. My research has been recognized with Best Student Paper Award in Research Track in KDD'22 and Outstanding Student Paper Highlight Award in AISTATS'24.

I have been lucky to be advised by Rahul Mazumder during my PhD at MIT. I was an Intern (Summer'20) / part-time Student Researcher (Spring'21) at Google Research (NY). Before that, I did my Masters at MIT (SM'18), and Bachelors at LUMS (BS'14, Gold Medalist), where I worked on power management solutions for power electronics. My CV is available here.