About
Hello! I’m a ML Researcher at MatX, working on training efficient LLM architectures (attention, MoEs, quantization) co-designed for our custom chip. Previously, I was a Research Scientist at Meta in the AI recommendation system co-design group, where I worked on designing efficient foundation models and used quantization/sparsity techniques to further accelerate training and inference performance. I graduated with a PhD from Cornell University, surrounded by brilliant folks at the Computer Systems Lab (CSL). I worked with Prof. Adrian Sampson in the CAPRA research group. My research spanned efficient machine learning, vector architectures and compilers.
I completed my Bachelor’s and Master’s degree in ECE from IIT Bombay in 2018. I worked with Prof. Sachin Patkar on Accelerating Sparse Matrix Solvers on FPGA for my thesis.
My research interests have varied over the years from Semiconductor devices (InAs Nanowire FET Modelling) to Machine learning and hardware-software co-design. Here is my CV.
Outside academia, I enjoy playing tennis, pickle ball, hiking, running and weightlifting.
Updates
- I joined MatX as an ML Researcher in Mountain View, CA. [May, 2026]
- I am serving on the Program Committee for ISCA 2025, and ASPLOS 2026. [Jan, 2025]
- I joined Meta as a Research Scientist in Menlo Park, CA. [July, 2024]
- I defended my PhD thesis, Software-Oriented Hardware Prefetching and Vector Execution. [June, 2024]
- I am continuing working at Google as a Student Researcher, with Akanksha Jain and Snehasish Kumar. [Sep, 2022]
- I am serving on the Artifact evaluation committee for ASPLOS, 2023. [Sep, 2022]
- I’ll be in Seattle visiting CSE @ University of Washington for the academic year 2022-23. [Aug, 2022]
- I did a summer internship in the Platforms and Infrastructure Performance team at Google. [May-Aug, 2022]
- Our paper on RISC-V vector compiler performance was published in the IEEE Micro special issue journal. [June, 2022]
- I passed my thesis proposal (A-exam). [May, 2022]
- I worked in the Machine Learning and Optimization group at Microsoft Research where we developed algorithms to accelerate sparse ML models on commodity hardware [May-Aug, 2021]
- I interned at Intel Labs, Santa Clara in the Parallel Computing group. I implemented high performance algorithm for sparse computation for Intel’s breakthrough architecture for a research project [May-Aug, 2019]