Notes drafted in seconds, signed by the worker
Workers summarise in their own words. CareBuddie returns a structured, outcome-linked note in seconds, tied to the participant's active NDIS goals. They read it, change what is wrong, and sign. The signature is theirs, not the model's.
Request early accessProgress-note editor. An AI-drafted, unsigned note tagged to two NDIS goals, ready for the worker to edit and sign.
The validated draft is stored alongside the signed note, so an auditor can see both.
Every note points at a goal someone is being funded for
A note that does not reference an outcome does not help at plan review. The draft is built against the participant's active goals and outcome domains.
- Active goals only · The draft references the goals that are live on the current plan.
- Outcome domains · Each note is tagged to the domain it evidences.
- Visible on the note · Goal tags sit on the note itself, so the link is there when someone reads it.
Note list. Each note carries the goals it evidences.
Once it is signed, it does not change
A note a provider can quietly edit after the fact is worth nothing in an audit. Signing freezes it. A correction is a new addendum row, with its own author and timestamp.
- Frozen on signature · The signed note is immutable from the moment it is signed.
- Addendum, not edit · Corrections are separate rows. The original stays visible.
- Draft kept alongside · The AI draft is stored with the signed note, so both are auditable.
A signed note with one addendum. The original is unchanged and the AI draft is retained.
AI never stands between a worker and the end of their shift
If the model is slow, wrong or unavailable, the worker writes the note themselves and signs it. Nothing about close-out depends on a third party being up.
- Always writable · Manual entry is a first-class path, not a degraded one.
- Never blocking · An AI failure falls through to manual entry.
- Rate limited · One generation per worker per minute, so cost stays predictable.
The manual path. Same structure and same signature, with no model involved.
Any questions?
Who signs the note?
The support worker. The model produces a draft; the worker reads it, edits it and signs it. The signature is always a person's.
What happens if the AI is unavailable?
The worker writes the note manually. AI failure falls through to manual entry and never blocks shift close-out.
Can a signed note be edited?
No. Signing freezes the note. Corrections are separate addendum rows with their own author and timestamp.
Is the AI draft kept?
Yes. The validated draft is stored alongside the signed note so an auditor can compare them.
Is participant data sent to a third party?
Generating a draft sends shift content to the model provider. What may leave the tenant is set out in a written decision record rather than decided feature by feature.
Discover more of the platform
Run your NDIS operations without the chaos.
Early access is by invitation. Your trial starts the day your invite lands. No credit card, no setup fee, and your data stays in Sydney.