By
@Praknoor
I built a tiered, cost-optimized AI agent for the AMD Lablab Hackathon 2 that maximizes token efficiency and accuracy by routing tasks through zero-cost parsers, local gemma checks, and a smart Fireworks escalation layer.
17h 31m 53s logged
yoo guys here is my agent for cost optimization, it was for lab lab amd hack 2 rn i am 22th in te leaderboard, but yeah let-s hope my model goes 1st 😭😭😭😭….so the neat thigh if this agent of mine is that it can run also on a 2-core 64-bit CPU, 4 GB RAM, 3 GB free disk * i mean these where the requirments, but yeah gng, my next step is to make it the next opencode rn the accuracy is 89% based onn the metric of the hack
