Agilance — AI Triage
An AI triage platform that helps patients describe chest pain by voice or text and gives clinicians structured, prioritized information before the visit.
What needed solving
Cardiovascular symptoms are difficult to structure before clinician review. Patients describe chest pain in their own words, often incompletely, and clinicians spend scarce time reconstructing what matters.
Care teams also had no reliable way to see, across many incoming patients, who needed attention first.
What made it hard
The experience had to work for anxious patients with no account and no training, by voice as well as text, in a few minutes.
Every output had to be clinically structured and reviewable. The assistant supports clinicians; it never replaces their judgment, so the information it produces has to be transparent and easy to verify.
What we built
Patients talk to an AI assistant that asks clinically structured questions about their symptoms by voice or text. As the conversation unfolds, the platform organizes symptoms, evaluates relevant risk factors, and turns the dialogue into structured clinical information.
A clinician dashboard brings those results together: patient histories, symptom summaries, risk visualization, and workflows that surface who needs attention first.
How it looks and feels
The patient side is deliberately calm: one question at a time, a visible listening state, and clear controls for mute, end, and audio. Reassurance is built into the copy, from the three-minute promise to the privacy note under the primary button.
The clinician side is dense by design. It favors scannable summaries, risk indicators, and consistent layouts so a care team can move through patients quickly.
How it works
The product combines a Next.js and TypeScript frontend with a FastAPI backend and a PostgreSQL database, with AI and voice infrastructure handling the conversational assistant and transcription.
Structured clinical output is generated server-side and stored in a form the dashboard can query, so patient conversations become data clinicians can act on rather than transcripts to read.
Where it landed
A production platform that takes a patient from an unstructured description of chest pain to a structured clinical picture in minutes, and gives clinicians a prioritized view of incoming patients.
For dev&dash, it demonstrates the full arc the studio is built around: product strategy, UX and UI, frontend, backend, AI integration, and deployment, delivered as one continuous effort.
From idea to production, we design and build the whole thing.
Start a Project →