"ROCmPilot team on AMD AI League Hackathon"
Team Idea
Building ROCmPilot, an autonomous AI inference optimization agent for AMD ROCm. It benchmarks LLM deployments, analyzes throughput, latency, TTFT, and memory usage, and automatically experiments with different configurations to find the best setup under user-defined performance constraints. I’m currently building solo and focusing on AMD Developer Cloud, ROCm, vLLM, and AI agent orchestration.
.png&w=256&q=75)
Submission
Submission draft in progress