About

I spent my first career in an airline cockpit, where nobody is asked to remember anything that matters: there is a written checklist, a second person answering it, and a recorder running the whole time. Then I started keeping track of appointments for myself and the people closest to me, and found the opposite arrangement. The important part gets said once, quickly, near the end, by someone who is already thinking about the next room. Nobody is being careless there either. It is a system that simply never installed a recorder.

So I built one, for myself and my family first. Building it is my second career, and PatientScribe is independent: no investors, no growth targets, nothing it has to become other than useful.

It is free. There is no investor waiting to be paid back, which is the usual reason a health app has to start extracting something from the people using it. Costs are low by design. This site is static files on a content network with nothing running behind it. On a current iPhone your recording is transcribed on the phone itself, so the expensive part often costs nothing at all. What is left, transcription on the paths that need it and the summary work, runs in cents per visit.

That math has limits. If PatientScribe ends up on a great many phones, those cents per visit add up, and a nominal fee may become necessary to keep it running. If that day comes, the fee will be sized to cover what the app costs, and the promise below will not change.

Your care recommendations are never monetized, at all. That is the promise I care about most. There is no referral arrangement behind this app, no pharmacy or clinic or discount program paying to be the thing it suggests, and nothing routing you toward whichever provider pays better. Your information is not sold or licensed either, identifiable or not, which the privacy policy says in more binding language than this page can. Keeping that promise is not hard. The moment the app steers your care for money it stops being the thing I set out to build.

PatientScribe is a record of what was said, never medical advice. It does not diagnose and it does not recommend, and nothing it writes should change what you do before you have talked with the person who examined you.

And the models that turn audio into transcripts and summaries make mistakes. They can mishear a word, drop a qualifier, or write a dose with more confidence than the room said it. That is why the audio and the transcript sit beside every summary: check the sentence against what was actually said, and when something reads wrong, trust the room, not the app.

Part of the work happens on servers rather than on your phone. Exactly what they receive, what they hold while your summary is being made, and when it is deleted is inventoried on Data & Privacy, in plain language.

Questions and reports come to me directly, and they are answered by the person who wrote the code. If that sounds like the kind of app you want in the exam room, the Founding Circle is where it starts.

D.