Umang Pandey#
Umang Pandey is an Applied Research Scientist in AMD’s AIG Models Group. His work focuses on AI-driven GPU kernel optimization, including evolutionary-search systems, reinforcement-learning pipelines, and large-scale synthetic-data generation for training kernel-generation models. At AMD, he contributes to building next-generation AI-for-Code models that optimize Triton and low-level GPU kernels.
Before joining AMD, Umang was a Research Fellow at Microsoft Research (MSR) in the PROSE group, where he worked on program synthesis, Excel automation, and ML-based code reasoning. He holds a bachelor’s degree in Mechanical Engineering from IIT Kanpur.