Five tracks, published before registrations close. You choose one at kickoff — and you do not need an idea to register.
The tracks below are categories, not assignments. Build whatever you want, then file it under whichever track fits closest. Nobody has ever been marked down for an idea that wasn't on this page.
The bullet points inside each track are there to unstick people who genuinely have no idea on day one. They are starting points. Original ideas score higher than ones taken from this list — the rubric rewards originality directly.
Problems you personally have right now. This track consistently produces the strongest first-year projects, because you are the user.
Small, respectful tools. Nothing that diagnoses, prescribes, or claims to replace a professional — judges mark that down heavily.
Local problems beat global ones here. A tool for your street will score better than a dashboard about world emissions.
Popular and therefore competitive. A wrapper around a chat API scores poorly; a specific workflow made genuinely faster scores well.
Anything that does not fit above. Games, developer tools, art, hardware simulations — if you can defend why it matters, it belongs here.
Pick the problem you have personally been annoyed by in the last month. That single heuristic produces better first hackathon projects than any brainstorming technique, because you already know what the solution needs to do and you can tell instantly whether it works.
Then make it smaller than feels right. A team that ships one screen that genuinely works beats a team that demos five screens held together with placeholder data — every single time, on every rubric line.
At kickoff, not at registration. Round 1 asks for the idea — no code — so you have until the round opens to make up your mind, and you can change track between registering and starting.
Entry closes before the idea does. Registration takes about two minutes, and you can pick your track later.