Joined AMD as a PhD Research Associate, working on concurrent GPU kernel execution and fine-grained dispatch strategies for LLM inference in vLLM.