HIPAA-Aware Healthcare Intake + Triage Bot
A patient-facing chat that collects symptoms structured for clinical review, routes to the right provider type, books the appointment, and writes an encrypted note to your EHR — with a full audit log.
A live AI receptionist picks up, books appointments, and answers common questions. It runs on a real phone line right now.
How it works
1. Intake with guardrails
Claude-guided symptom flow emits structured JSON while PHI-redaction guardrails scrub identifiers from any outbound prompt logs.
2. Route and book
A provider-type router classifies PCP vs specialist vs urgent and books into the correct calendar — integrated with Google / Athena / Epic mock APIs.
3. Encrypted EHR write + audit
Structured note is encrypted at rest, written to the mock EHR, and every PHI touch is logged with user, timestamp, and purpose.
60-second Loom
60-second flow: patient describes symptoms → structured capture → routing logic → appointment booked → encrypted EHR write + audit-trail entry.
Get the full build spec PDF
One-page PDF with the architecture, the tech stack, the timeline (2–3 weeks), and the investment range ($9,990 – $25,000). Sent to your inbox in under a minute.
Who it’s for
Digital-health startups and clinics who need a HIPAA-defensible intake surface: BAA-ready hosting, encryption at rest and in transit, full PHI audit trail.
Tech under the hood
Frequently asked questions
Is this HIPAA compliant out of the box?
It’s HIPAA-aware: BAA-ready hosting on AWS Bedrock or Azure OpenAI, encryption at rest and in transit, per-patient RLS, PHI redaction on outbound prompt logs, and a full audit trail. A formal HIPAA certification requires your BAAs and policy sign-off.
Which EHRs does it integrate with?
Athena, Epic (via FHIR + Orca), eClinicalWorks, DrChrono, and any EHR with a FHIR R4 or HL7 v2 interface. The demo ships a mock EHR writer you swap for the real endpoint.
What about non-English patients?
Claude handles 30+ languages natively. Intake flows auto-detect and adapt. Translated clinical notes include a ‘translated by AI, human-reviewed’ attestation for EHR documentation.
How do you handle red-flag symptoms?
Red-flag keywords (chest pain with radiation, stroke symptoms, suicidal ideation, pediatric fever >104°F) immediately route to a live agent queue and surface emergency-dial UI. Configurable per specialty.
Want one of these in your business?
We’ll scope, build, and hand off a production-ready system in 3–4 weeks.