About
Oncology data holds enormous promise for AI, but only if teams can tell when a dataset is actually fit for the question being asked, and can explain why.
Join The Open Accelerator, OncoSys AI, and Northeastern for a four-day hackathon to build practical tools that make oncology data safer, clearer, and more useful for responsible AI.
Schedule
Wednesday, November 11: Virtual kickoffThursday–Friday, November 12–13: Virtual Build daysSaturday, November 14, 10:00 AM–5:00 PM: In-person finals at the Cabral Center, Northeastern University, Boston
Challenge tracksTrack 1: Cohort Firewall
Build a tool that decides whether an oncology dataset is ready for a proposed AI question. It returns one of four verdicts: Ready, Conditional, Blocked, or Insufficient Evidence.
Track 2: Evidence Navigator
Turn complex oncology data-readiness findings into clear evidence, side-by-side comparisons, and concrete next actions.
Before you register
Team size: 3–5Bring a working prototype to finals. Slides alone won’t do it.Use only synthetic, openly licensed, or event-approved non-sensitive dataNo patient data, and no tools that make clinical decisions
Who should joinDevelopers, data scientists, designers, researchers, and anyone working where healthcare meets AI. Cohort Firewall suits builders who like rules, logic, and data. Evidence Navigator suits people who can make complex findings easy to understand.
Organized by
Northeastern University · OncoSys AI · The Open Accelerator · Red Hat
Listed from Luma