Clinicians That Code

The directory of clinicians that build software

Doctors, nurses, pharmacists and other clinically trained people who do not just use software; they ship it. Every entry is verified: a real clinician, a real, personally built artefact.

70 verified clinician-builders: search and filter by role, how they build, and where.

Pulse: what the ecosystem surfaced recently, last update 2026-09-28. Library: the monthly archive of everything before that.

Guides: practical answers, from the compliance lane you're in to which platforms will sign a BAA.

The Handover, our weekly five-minute briefing on what happened at the crossroads of medicine, code and regulation.

Why this exists: The 3am Nurse Test, on the bottleneck that just fell and what clinicians can build now.

The Handover

One email a week. Five minutes.

What happened at the crossroads of medicine, code and regulation, every Friday. No vendor marketing, no filler; the archive lives here.

About

Clinicians That Code tracks a specific wedge: clinically trained people who personally build and ship software; not clinicians that merely use AI tools, and not vendors selling to them. If a doctor vibe-codes a dosing calculator on a night shift, we want it on the map.

The directory is hand-curated and verification-gated; the pulse is fed by an automated monitoring pipeline that sweeps Reddit, X, LinkedIn and the health-tech press three times a week.

Curated by Bart Collet. Part of the Care & Code family: the community event for clinicians that build, in Antwerp on 3 October 2026.

Built something yourself, or know a clinician who did? Suggest an entry.

What is this?

Which AI coding platforms sign a BAA? A handful of the model APIs and IDEs clinicians already use do, as of the last check; several widely used ones don't, and "no BAA" isn't a verdict on quality, just on whether real patient data can touch that vendor. See the BAA fact table for the full, vendor-sourced breakdown.

Is vibe coding HIPAA compliant? It depends on what the build touches, not what it is: prototyping on synthetic data is unconditionally safe, real patient data or a live clinical workflow pulls in data protection and access controls, and influencing real diagnosis or treatment is device territory. See Build safely for the three lanes.

How do I get patient data out of the EHR as a clinician? Two genuinely different routes: your own right of access to your patients' records, or an EHR vendor's developer/FHIR programme, each gated by legal, technical and institutional layers that a "FHIR-enabled" badge alone doesn't clear. See What data can you actually get out of the EHR for the Epic/Oracle Health/eClinicalWorks breakdown.