AI across Rostery
Eighteen places the software uses AI, grouped by where you meet them. Every one of them drafts, suggests or summarises — a person still decides and a person is still responsible. Nothing here acts on its own.
19 guides · 108 steps · 2 screenshots
AI Client Brief
A short brief on a participant before you work with them: medical history, risks and support needs, pulled from their own record.
Write a progress note
Turning rough words into a proper progress note, and tidying one that is already written.
Care plan and goals
A first draft of a care plan, written by AI from what is already on the participant’s record. You review it before anything is saved, and nothing reaches the record until you import it. The participant’s NDIS goal register is kept separately, on the Goals tab.
Watching for deterioration
An AI scan across your active participants for signs things are getting worse: more incidents than the month before, cancelled shifts, missed medication doses, medication changes and incidents still open. It tells you who to look at. It does not diagnose anyone.
Suggested workers for a shift
Ranking workers for a shift by how well they actually fit it.
AI Auto-Roster
Building a draft week of shifts in one go, from patterns, availability and qualifications.
Burnout scan
An AI read of each support worker’s last 30 days: hours, completed shifts, cancellations and absences, leave, incidents and shifts longer than 10 hours. It flags who may be heading for burnout while you can still change the roster.
Is this incident NDIS reportable?
A second opinion on whether the Commission has to be told.
Incident autofill and root cause
AI help with incidents. Autofill drafts the structured fields of a report from the account you type, and root cause analysis suggests what may be behind incidents that keep repeating. Both are drafts for you to check, and nothing is saved until you save it.
Incident patterns
A banner at the top of Incidents that groups the last 30 days to show a recurring problem no single report reveals: the same kind of incident, the same participant or the same place, again and again.
Reading an uploaded document
Pulling the details off a certificate so you do not type them.
Checking an invoice
An AI check of one of your invoices before it is claimed. It compares each line with the shift and timesheet behind it, looking for the problems that get a claim rejected or queried. It is advisory: the claim rules still decide.
Funding running out
The Fund Depletion Forecast report works out when each participant’s funding will run out at the rate it has been spent, and flags the funds most at risk. An AI Forecast can also weigh up the shifts already booked.
Summarising a report
A short, plain-English summary of what a report is showing, written by AI from the report’s rows. It picks out the key numbers and the single most important action, if there is one. Check the figures before you quote them.
AI Receptionist
Answering your phone, having a normal conversation, and capturing the enquiry as a lead with a transcript.
Website chatbot
The chat assistant on your organisation’s website. It answers visitors’ questions from the knowledge you give your AI receptionist, and saves each enquiry as a web-chat lead for your team to follow up.
Drafted reminder emails
AI-drafted reminder emails for staff whose documents expire in the next 30 days. It writes one short, professional email per person, naming each document and its date. Rostery does not send them: you copy each one into your own email.
Connect an AI assistant to Rostery
Letting your own AI assistant (such as ChatGPT) read and, if you allow it, write operational data — so you can ask it questions in plain language.
What the AI will not do
Worth knowing before you rely on any of the above.