Your notes are your receipts. We make sure they hold.

That depends on whose account the notes were written in, and it is the single thing most worth getting right. Records created in a support worker’s personal account belong to that worker: your organisation cannot access them, cannot export them, and cannot produce them when somebody asks for them. When the worker resigns, they go too, and support work turns over at roughly a quarter of the workforce a year. Records relating to the payment and receipt of NDIS amounts must be retained for seven years, with a civil penalty for failing to. Twenty free individual accounts is twenty disconnected note writers and no organisational record. An organisation account holds every worker’s notes in one place, under your control, attached to the claims they substantiate. (Retention rules are the Agency’s, not ours. Confirm the current position with them.)

It scores a progress note against the specific claim it substantiates, rather than against how well it reads. NDIA Provider Payment Assurance guidance sets out what a provider is expected to be able to produce when a claim is reviewed: for complex supports, typically both a log or roster and a case note, evidencing the quantity of support and the type of support delivered, alongside a service agreement. Our six checks are an implementation of that published requirement, not our own opinion about good writing: does the note evidence the support item claimed, do the hours match the hours claimed, is the support linked to a goal in the participant’s plan, is the language objective and specific, are the required elements present, and is there anything in the note contradicting the claim. You get a score, the specific gaps, and what to do about each one. A check we cannot judge from the data available is excluded rather than failed. The score is a heuristic, not a regulatory determination.

It is the most common quality problem in progress notes, and it is not a training problem you can fix by asking people to try harder. A large share of the disability workforce speaks English as an additional language, and a worker who delivered excellent support can still produce three rushed lines that would not substantiate the claim behind them. So ShiftNote lets the worker describe the shift in the language they actually think in, spoken or typed, across 93 languages, and writes the note up in professional Australian English in the same six-section structure every other note uses. Three things are worth being precise about. It translates and tidies the wording; it does not add detail, confidence or clinical language the worker never gave. Anything ambiguous is flagged for the worker to confirm rather than guessed at, because a confident mistranslation in a care record is worse than an open question. And the worker’s original words are stored with the note, so the English version can always be checked against what was actually said. Speaking out loud is supported in most of those languages and depends on the worker’s device; where it is not available they type, and the note comes out the same. As with every note here, it is a draft the worker reviews and confirms, and it is not a certified translation.

Claim validation asks “is this claim formatted correctly?”: price limits, item codes, budgets, missing fields. Several systems do that well. We ask a different question: “if the NDIA pulls this claim in three years, does the paperwork behind it hold?” A claim can be perfectly formatted and still be unsubstantiated because the note behind it describes something else, or does not exist.

No - and you should be wary of any tool that claims it does. ShiftNote produces NDIS-aligned progress-note drafts structured for review, and an evidence score that is a heuristic, not a regulatory determination. Your team remains responsible for checking, editing, and approving every note, and for meeting your provider and NDIS obligations. What we stand behind is our own work, not the regulator’s verdict.

No, and we are not trying to be. ShiftNote is a layer that sits alongside the care-management system you already run. There is nothing to migrate and nothing to switch off. To be clear about what that means today: we have no API integrations with any of them. Notes are copied across, and claim and roster data comes in by CSV import or manual entry.

Provider plans are metered on how much you claim per month, not on how many workers you have. Every paid plan includes unlimited workers. Team, up to $150k/mo claimed, $299/month; Provider, $150k–$600k/mo claimed, $899/month; Assurance, over $600k/mo claimed, from $2,400/month. Annual is two months free across all three. Every provider is free until 1 December 2026; the prices apply from that date and we will give written notice before any organisation is charged. Individual support workers are free permanently, which is not tied to that date.

Because seats measure how many people you employ and claims measure how much revenue is exposed, and those are not the same number. A 2:1 SIL shift bills roughly double a 1:1 community access shift per worker-hour and carries far more scrutiny when it is reviewed. Two providers with identical headcount can be carrying very different amounts of risk, and seat pricing cannot see the difference. It also means a casual doing four shifts a month does not cost you the same as a full-timer. You tell us your approximate monthly claim value at signup and it confirms itself from the claim data you have already imported; there is no separate report to file and no audit of your finances by us.

Yes, permanently, with no card and nothing to cancel. This is not an early-access price that expires, and it does not end on 1 December 2026 along with the provider pricing; the individual plan stays free. You get progress notes, incident reports, voice input, full history and PDF export, with a fair-use limit of 50 generated notes a month (a full-time roster is 20–25 shifts). We charge organisations, not workers. Turning rough notes into a structured draft is not the hard part any more, and charging one person for it just pushes teams into opening separate accounts.

They can, and some will, and we would rather they used ShiftNote than not. But 20 individual accounts give a provider 20 disconnected note writers: no exposure number across the team, no claim linkage, no 90-day clock, no evidence pack, and no admin visibility. Those need every worker’s notes in one place plus your claim data, so they cannot be assembled from personal accounts at any headcount. And see the first answer above: those records belong to the workers, not to you.

Send us three months of claims and the notes behind them, and we produce a report showing how much you have claimed with weak or missing evidence behind it, broken down by worker, support item and participant, plus which claims are still inside the 90-day window. It is $1,950 as a one-off, credited in full against your first 3 months on any plan, and free for the next 2 design partners, until 31 October 2026. You keep the report whether or not you go further. There is also a free self-serve version at /exposure-audit that runs entirely in your browser; it can only show claim-window exposure from dates, not evidence exposure, because that needs the notes.

Yes. A provider admin creates an organisation, sets a note template, and shares an invite code or link. Workers join in a tap, and the admin sees aggregate usage from one simple dashboard.

Never. Every output is clearly a draft. Before a note can be copied, downloaded, or sent, the worker confirms a short review checklist. The worker stays the author and remains responsible for the final note.

If the details mention things like a fall, injury, medication error or behaviour of concern, ShiftNote shows a neutral reminder to follow your provider’s incident management and reporting process. It does not diagnose anything or decide whether something is legally reportable - that stays with you and your provider.

No. We never send note content, participant names, reference numbers, diagnoses, medications or incident details to analytics. We track only privacy-safe usage events such as note created, note copied, and template updated.

In Australia. Our database runs in the ap-southeast-2 (Sydney) region, migrated 2026-08-06. Your notes are never used to train any AI model.

Never. Notes are processed by Anthropic’s API to generate a draft and are not used to train models, by us or by them. Every note stays yours.

Yes. Provider admins can set the preferred tone, structure (paragraph, bullet, SOAP-like or custom), required fields, and a custom instruction. Those preferences shape every draft your team generates.