I am a Senior Researcher at Microsoft Research in Redmond, where I work on designing system optimizations for LLM inference and RL post-training frameworks. I am excited about achieving high GPU utilization at scale, and like to work on and co-design different angles of this problem, such as scheduling, network communication, memory optimization, and resource allocation.
I obtained my Ph.D. from the Computer Science department at the University of Texas at Austin (UT Austin) where I was advised by Vijay Chidambaram and worked closely with Philipp KrĂ€henbĂŒhl. My Ph.D. dissertation (linked here) provided automated solutions for efficient network utilization and memory capacity consumption in ML systems. Prior to that, I received my Bachelor's degree in Computer Science and Engineering from the Indian Institute of Technology (IIT) Roorkee.
| Mar 2026 | MSCCL++ received honourable mention for Best Paper Award at ASPLOS. |
| Jan 2026 | MSCCL++ accepted at ASPLOS'26! |
| Oct 2025 | Gave a guest lecture at UW on optimizing collective communication libraries. |
| Mar 2025 | Splitwise chosen in IEEE Micro Top Picks 2025. |
| Apr 2024 | Gave a guest lecture at Cornell on optimizing collective communication in large-scale ML systems. |
| Mar 2024 | Splitwise accepted at ISCA'24! |
| Aug 2023 | Started as a Senior Researcher at Microsoft Research, Redmond! |