P-02 — Case study
Patient intake with AI triage
An intake flow that asks the right follow-ups and hands clinicians a summary, not a form dump.
- Client
- Halden Health
- Sector
- Healthcare · AI
- Year
- 2025
- Status
- live
- Modules
- AI & AutomationSoftware & SaaSUI/UX & Product Design
01
The brief
Patients filled a twelve-page form; clinicians read none of it. Intake took twenty minutes of appointment time and still missed the thing that mattered.
02
The build
A conversational intake that adapts its questions, grounded in the clinic's own protocols through retrieval, with a clinician-facing summary that cites every answer. Every model output is evaluated against a reviewed set before release.
03
The result
Intake moved out of the appointment. Clinicians start with a one-page summary they trust, and the practice recovered the time it used to spend reading forms.
Outcomes
1 page
clinician summary, every answer cited
-18 min
per appointment
100%
of outputs evaluated pre-release
FHIR
native records integration
- RAG
- LLM agents
- FHIR
- Next.js
Next on the board · P-03
Headless storefrontMara Roasters · E-commerceReady when you are.
Tell us what you're trying to build. We'll come back with how we'd build it — usually within a day.