Software, not slide decks. Here's what our team is actually shipping.
The problem: Traditional sickle cell screening relies on lab workflows that aren't always fast or accessible, especially outside major hospital systems. We're building an AI model aimed at faster, more accessible screening.
Who's building it: A mixed team of premed members (defining the clinical problem and evaluation criteria) and engineers (building and training the model), working together rather than handing requirements over a wall.
Status: Actively in development. We're not publishing performance numbers or a public demo yet, on purpose, until we're confident the results are real and reproducible.
Sickle cell detection is our first build, not our last. Once it ships, we're planning to take on additional AI-in-medicine projects nominated and scoped by our premed members. If you have an idea for one, that's exactly the kind of thing we want to hear about.
Every project starts on the premed side, not the engineering side. A member identifies a real clinical gap, defines what success would actually look like for the people affected by it, and only then do we start building. The model serves the problem, not the other way around.
We're also deliberately slow about claiming results. A premed-built AI tool that overstates its accuracy does more harm than good in a medical context, so nothing leaves "in development" status until it's been validated and reviewed by more than one person on the team.
We're a small team and we need both sides: engineers to build, and premed members to keep us honest about the clinical problem.
Join the Build Team