Solo builder competing in the AMD AI League. I’m an MS Engineering Science (AI) student at the University at Buffalo, building a reusable, token-efficient agent framework on AMD ROCm GPUs. It combines a plan-act-verify loop, DOM-first browsing with a vision fallback, and evidence-backed answers. I’ll apply it across the Matches (web retrieval, code repair, unknown-game agents) and publish it open source with write-ups for the community.
.png&w=256&q=75)
AI Developer Intern