A ValiDATAthon is a structured, collaborative, hackathon-style, single-day workshop co-hosted with our partner institutions.
What is a ValiDATAthon?
During a ValiDATAthon, clinicians independently review blinded AI responses to simulated contextual clinical scenarios, scoring them against defined criteria such as clinical soundness and safety.
Because the scenarios are hypothetical, no patient data is used, and there is no privacy risk.
What You Gain From Taking Part
Benefits for Partner Institutions
National recognition as an official ValiDATAthon partner plus a spotlight on our website
Benefits for Participating Clinicians
Certification in AI Evaluation
How a ValiDATAthon session works
Stage 01 Pre-ValiDATAthon Survey & Consent
Before the session begins, participants complete a short baseline survey and review informed consent, giving a starting point to compare against once the session is done.
Stage 02 AI Knowledge Session
Clinicians take an interactive crash course on how a frontier medical AI model is actually built: the data it learns from, how it is trained, and where it can go wrong, so they gain a grounded understanding of how AI works before they evaluate it.
Stage 03 Clinical Expert Evaluation
In the hands-on core of the session, clinicians review realistic but hypothetical scenarios alongside the AI's responses, then evaluate each one across dimensions like accuracy, safety, and clinical appropriateness.
No real patient data usedStage 04 Post-ValiDATAthon Survey & Feedback
Participants complete a closing survey that measures the shift in their confidence in medical AI and gathers structured feedback on the experience.
