Hacktober Hack
A free two-hour hackathon where teams make a flaky AI agent shippable, judged on hidden eval runs.
A hackathon on building quality agents, co-hosted by Failproof AI and Quad AI.
We hand every team the same AI agent. It works... sometimes. You have under two hours to make it one you would ship.
Same agent, same model, same tools for everyone. Change anything around the model: prompts, tool definitions, validation, retries, guardrails and policies. Then we run a hidden eval suite against every submission, three times per task. A task only counts if it passes all three. How the evening runs 4:30 pm: doors open, form teams 4:45 pm: kickoff. Meet the agent, the public eval suite and the baseline score 5:00 pm: build 6:50 pm: code freeze. The hidden suite runs while teams demo 7:30 pm: results and wrap Judging Reliability (80 points): the share of hidden tasks your agent passes on all three runs, inside the call, time and token limits
Your evals (20 points): up to ten eval cases you add, each counted when the baseline fails it and your agent passes it every time Ties go to the lower cost per task Bring your laptop, your charger and a GitHub account. Come solo or as a team of up to three. We provide the agent, a model key for your team and the public eval suite.
It's Hacktoberfest, so the agent is open source. After the event, open your fix as a pull request. The best fixes and evals get merged.
Free. Registration is subject to host approval. The exact address in HSR Layout is shared once you're in. Also listed on MLH: https://events.mlh.com/events/14866-hacktoberfest-meetup-be… Venue: Quad AI HSR Campus, 27th Main Rd, beside RBIH, 1st Sector, HSR Layout, Bengaluru 560102. Free. Registration is subject to host approval.
