Sonus Medic: from appointment to finished record

The problem

Clinical admin is relentless. After every appointment a clinician re-types what just happened into forms and reports, then files it into the patient record — minutes of documentation stacked onto every consultation, adding up to hours a week that could have gone to care.

What we built

Sonus Medic records the consultation, transcribes it, and auto-fills the clinic's own forms and Word reports — not a generic template — straight into the patient record. The clinician reviews and corrects the draft before saving, so each appointment ends as a complete, EHR-linked file. The result is far less admin per patient, with the clinician always the final author.

Where AI earns its place

Transcription and structured drafting are exactly where models shine, and exactly where a human must stay in the loop. The AI proposes; the clinician disposes — reviewing and correcting every record before it is saved. Nothing reaches the patient file without a professional's sign-off.

Approach & stack

We built the capture-to-record pipeline end to end with AVG/GDPR-proof handling throughout — a hard requirement for Dutch clinical data — on a modern TypeScript foundation with containerised, EU-hosted deployment and clinical-grade transcription tooling. Because it writes into the clinic's existing forms and patient record rather than replacing them, it fits the way each practice already works instead of forcing a new one.

Outcome

Sonus Medic is live with practising clinicians. Every appointment ends as a complete, EHR-linked record instead of a pile of after-hours paperwork — the documentation happens as the consultation happens, with the clinician reviewing and signing off before anything is saved. Less admin per patient, and AVG/GDPR-proof handling throughout.