Ravi Gupta#
Ravi is an algorithms and computing researcher in Distributed Computing who analyzes and optimizes workloads at scale using large supercomputing clusters. His experience spans large-scale systems, algorithms, and performance tuning and optimization at scale. He completed his Master’s thesis at Purdue University, funded by LLNL.
Since then, for a large part of his career, he has worked with top national laboratories such as Lawrence Livermore National Laboratory (LLNL) and Argonne National Laboratory (ANL).
Most recently, he works on distributed disaggregated inference involving llm-d, vLLM, and SGLang, and related nuances in different parallelism types such as expert parallelism.