Channel K03 · 2026
ZorgNotitie
Live demo · built on my own initiative
The AI proposes the care note, the care worker decides (demo with synthetic data).
Try live (in Dutch)
In short
- Built for
- Care teams (demo with synthetic data)
- The problem
- After every visit, time goes into record-keeping instead of into the client
- What it delivers
- Speak instead of typing: the AI makes a draft, the care worker checks it and saves it themselves. Try it live.
Overview
After a visit, a care worker dictates a short note. The AI turns it into an organised draft: care given, deviations, mood and behaviour. That is where the AI stops: the care worker checks, adjusts and only then saves it themselves. No risk scores, no diagnoses, no automatic reports.
Challenge
After every visit, care workers spend time on structured record-keeping: the so-called "five-minute registration". Organisations differ in how far along they are, but reporting deviations still causes friction for many care workers, in a sector with 68,500 open vacancies.
What it does
- 01Dictate
After the visit, the care worker dictates a short note, simply in the browser
- 02Draft
The AI turns the note into an organised draft, but saves nothing
- 03Check
The care worker sees the note and the draft side by side and adjusts where needed
- 04Save
Only after explicit approval do points of attention appear on the team overview
Result
- The AI proposes, the care worker decides: nothing is saved automatically
- No risk scores, diagnoses or automatic reports, only fixed points of attention
- For each care moment you can see what the AI proposed and what was saved
- Demo with synthetic data: not a medical device, no clinical decisions
Technology
Whisper transcribes the voice note, and an LLM turns it into a structured draft next to the raw transcript. The care worker checks, corrects and saves explicitly. Only then do deterministic points of attention (e.g. "mood different from normal") become visible to the team, never before.