Most hospital AI pilots don't fail because the model didn't work — they fail because nobody planned for what happens after the demo. This session is an honest, no-slides post-mortem: facilitators and attendees dissect real pilots that stalled, separating what actually killed them — organisational friction (workflow misfit, unclear ownership, clinician buy-in, change fatigue, budget/procurement gaps) from technical causes (data quality, integration debt, model drift, validation gaps) — and get concrete about what would have unblocked them. The goal isn't a list of best practices; it's a candid working session on the specific decisions, timelines, and tradeoffs that separate a pilot that scales from one that quietly disappears.

This workshop is designed to move discussion into real-world action. Each session creates space for meaningful conversation, peer learning, and shared problem-solving, with a clear focus on what happens after the event. Participants will be encouraged to continue the dialogue beyond the room by sharing reflections, resources, and follow-up actions via a dedicated WhatsApp group. By building these connections, the Action Lab helps turn a one-off session into an ongoing community of practice.