ClearTriage is a medical intelligence layer that sits between a symptom and a decision. It does not diagnose—it structures uncertainty, surfaces relevant mechanisms, and points you toward the right level of care.
Open the Prototype →Most triage tools force a binary: emergency or not. Real medicine lives in the gradient. A headache can be tension, or it can be a sentinel bleed. The difference is in the pattern, the context, the mechanism.
ClearTriage was built to hold that complexity. It weighs severity across multiple axes—anatomy, physiology, time course, risk factors—and renders a structured report that respects the nuance of the human body.
Every output follows a rigorous schema: severity, facility type, mechanisms, next steps. No vague reassurance. No alarmism.
Research insights and biological mechanisms are surfaced alongside the triage decision. You see the reasoning, not just the result.
When appropriate, the tool finds real nearby facilities via OpenStreetMap—bridging analysis to action without friction.
ClearTriage emerged from a simple observation: people Google symptoms because they need structure, not because they want to self-diagnose. The problem is that search results are noisy, forums are biased, and most triage apps treat medicine like a flowchart.
This is a prototype built for the World Wide Vibes Hackathon. It demonstrates what happens when large language models are constrained by medical schema, grounded in real geolocation data, and designed with the restraint of a clinical interface. The full vision extends far beyond what you see here.