How projects are judged.
Judges include doctors, nurses, engineers and founders. We publish the rubric early, so you know exactly what counts before you start.
Five things judges look for
- 25%
Impact
Does it solve a real problem for a real person? Did you talk to someone who has it?
- 25%
Execution
Does it work? Can you show it live, not just in slides?
- 20%
Built with AI
How well did you direct and check the AI? Did it make the project better, not just faster?
- 15%
Design
Is it clear, kind and accessible? Would the person you built it for want to use it?
- 15%
Responsibility
Is it safe, private and honest about its limits?
From submission to the stage
Submit
By 11:30 AM on day two: your code, a short demo video if you like, and a one-page build log of how you used AI.
Expo
Science-fair style. At least three judges visit your table for about five minutes each.
Finals
The top teams present for three minutes on the main stage, then take questions.
Awards
Track winners, special awards and the Viva Prize are announced at closing.
Ground rules that keep it fair
- Start from scratch. Ideas and learning are fine ahead of time, code is not.
- Public or synthetic data only. No real patient data, ever.
- Say how you used AI. Disclosing it never costs you points; hiding it does.
- Projects that give health information must say when to see a professional.
- Respect licenses for any code, data or models you use.
Show us how you worked with AI
Every team turns in a one-page build log: what the AI did, where it was wrong and how you caught it. Judges read it alongside your code. Using AI well means directing it clearly and checking its work.