Daily AI check-in calls that keep your people in charge.
careplans ai is a Sydney company. We build voice check-ins for Australian aged and home care. A warm AI voice, always identified as AI, phones a client on any phone. Your care team hears about anything that matters and decides what happens next.
Good morning Doris, it is Mary from your home care team. How have you been since we last spoke?
Oh... not too bad, love. I had a bit of a fright in the shower this morning, actually. My foot slipped, and I only just caught the rail.
Oh Doris, that sounds frightening. Are you hurt at all? Did you hit your head?
No, no. Just shaken. But that floor gets so slippery, and I am a bit nervous about it now, to be honest.
I am so glad you are not hurt. Would you like me to ask your care team about making the shower a bit safer? And I can let your daughter know, too.
Oh, would you? That would be great. Thank you, Mary.
Every escalation has a named owner: your care team. Staff decide.
Illustrative example: a sample call with Doris, a made-up client. Mary says she is AI at the start of every call.
St Vincent's would start from a service already shaped by a Catholic provider's governance, security and ethics review.
As part of the trial, ethicists from the Plunkett Centre, which St Vincent's knows well, tested Mary and gave a written review. It covered how an AI voice speaks about itself, managing expectations, honouring a client who declines, and staff moral distress. Word choice became a formal ethics workstream.
The St Vincent’s mission is to express God’s love through the healing ministry of Jesus, with a commitment to people who are poor or vulnerable. Bringing Care Home asks for more care in homes by 2030. Your national AI Clinical Council is setting out to make AI ethical and effective. Catholic Health Australia asks that the common good and human judgement guide AI.
We built careplans ai on the same ground. The technology reaches into the home. People make the decisions.
Mary listens on the call. Kate, the coordination engine, turns what Mary heard into a short list of proposed steps. Every step comes from the Care Action Catalogue, the agreed list of actions a call can lead to, each with an urgency level, an owner and an escalation rule, built with Catholic Healthcare. Every example uses a made-up client.
“My foot slipped and I only just caught the rail.”Doris, on a check-in call
Treat it as a near-fall. Tell the care team today, ask for a home safety check, keep her daughter informed with her consent, and ring back tomorrow.
Tool Raises a mobility concern for the care team
Tool Sends a home safety request to the coordinator. Equipment may be funded under Support at Home assistive technology.
Tool Sends a Nonni update, only where Doris has agreed
Tool Schedules the next call
The coordinator sees one short summary with a named owner. The care record holds the near-fall, the request, the consent given and the booked follow-up.
“I only have a few of my heart tablets left.”Arthur, on a check-in call
Record which medicine is low and how many days remain. Ask the coordinator to arrange a repeat. Mary does not contact a pharmacy.
Tool Logs a medication check
Tool Sends a refill request to the coordinator
Tool Flags the list for a nurse or pharmacist to review
Tool Schedules the next call
The request waits with the coordinator, who decides. The care record shows the medicine, the date raised and the follow-up call.
“I missed my doctor yesterday and I have no way to get there.”Ruth, on a check-in call
Note the missed appointment, ask the coordinator to rebook it and to arrange a ride, and remind her the day before.
Tool Sends a booking request to the coordinator. Kate making the booking herself is in development.
Tool Sends a transport request to the coordinator. Arranging it with a supplier is in development.
Tool Gives a gentle reminder on the next call
Tool Adds a note to the care plan for the care team
Nothing is booked until a person approves it. The care record shows the missed visit, both requests and who approved them.
“I feel very alone since Frank died. I would like to talk to someone.”Maureen, on a check-in call
Listen, record it, and pass the request to a person. Mary gives no counselling and no faith content. Pastoral care is human-led.
Tool Scores isolation risk over recent weeks
Tool The independent safety monitor brings a person in straight away if needed
Tool Requests a chaplain. A person then contacts her. Switched on per service.
Tool Schedules the next call
A person responds, not the AI. The care record shows the request, who it went to and when they made contact.
“I got home yesterday. I am a bit wobbly and not sure what the new tablets are for.”Frank, on a check-in call
Schedule check-in calls across the first week, flag the medicine confusion and the unsteadiness, and summarise the discharge papers for the clinician.
Tool Reads and summarises the uploaded discharge summary for the clinician
Tool Asks a nurse or pharmacist to explain them to him
Tool Raises a mobility concern for the care team
Tool Books a series of check-in calls
The clinician has a summary and the care team owns the two flags. The care record shows each call across the week.
“Oh, I am fine. Nothing much to say.”Joan, on a check-in call
Over a week her calls have grown shorter and quieter than her usual. Note it early, ask a person to look, and add a gentle check-in.
Tool Scores isolation risk over 14 days
Tool Adds an observation to the care plan
Tool Asks Mary to run a structured wellbeing screen on the next call
Tool Schedules the next call
A change that nobody asked about reaches a person early. The care record shows the pattern and who looked at it.
Done automatically Kate does it and records it. Waiting for coordinator approval A person must approve before anything happens. Care team decides The team chooses what happens next. Steps marked In development are not yet running for clients.
Human-led. For pastoral and faith care, Mary notices and passes the request on. A person responds. The AI never provides faith content. This module is switched on per service.
St Vincent’s can review and tailor the catalogue to its own policies before any pilot.
The call is only the start. Around each client, the careplans family works together so that no single person has to carry it, and your people always decide.
A warm voice phones the client on any phone, says she is AI, and listens.
An independent AI safety monitor runs on every call.
The care team dashboard, escalations with a named owner, and compliance evidence written as calls happen.
Scheduling and follow-up calls, so nothing is left hanging.
In developmentBookings and supplier arrangements, such as transport.
See it in action in the examples above.
Your Council will ask what happens when the AI gets it wrong. Here is the plain answer. Mary does not diagnose or give clinical advice. She supports your clinicians and never replaces them.
Real escalation steps, contacts and response times are agreed with your care team before the first call.
These are the questions worth asking any provider of daily check-in calls. Here is how we answer each one.
| Question | Our answer |
|---|---|
Who decides what happens after a concern?Why it mattersClinical and ethical accountability must sit with your people. | Your care team. The call escalates to them, and an independent AI safety monitor runs on every call. |
What if it gets something wrong?Why it mattersEvery system makes mistakes. What matters is who catches them. | A person confirms before a request for care goes ahead, an independent safety monitor runs on every call, and the record shows what was heard. See the safeguards above. |
What security assurance is there?Why it mattersProcurement and your Council will ask for evidence, not assurances. | Aligned to ISO 27001:2022. We completed the vendor security review for Catholic Healthcare and will share our security documentation on request. |
How does it fit our systems?Why it mattersNew tools must not add double handling for busy coordinators. | Calls reach clients on their own phones and concerns go to your nominated contact with the call summary attached. We would agree any link to your client systems with your IT team before a pilot. We assume none. |
What will it cost?Why it mattersA pilot should have a known cost before it starts. | We price a pilot with you once the service, the number of clients and the outcomes are agreed, so the cost is clear before the first call. |
Does it work on the phones clients already have?Why it mattersApps and devices leave out the people who most need contact. | Any phone. No app. No device. |
Was it built for Australian aged care?Why it mattersEvidence must fit Support at Home and the strengthened standards. | Built in Sydney, with Care Minutes and quality indicator tools, and evidence written as calls happen. |
Does it respect your mission and faith life?Why it mattersSpiritual care is part of whole person care. | The pastoral module routes spiritual need to a human chaplain. The AI delivers no faith content. |
Can it notice change for each person?Why it mattersA general score can miss a quiet decline in one client. | In development: notices based on each client's own usual pattern. |
Are families kept in the loop?Why it mattersFamilies are partners in care and often the first to worry. | Post-call and weekly family summaries. |
Does it speak your clients' languages?Why it mattersCare in a first language builds trust. | Many languages, including Italian, Chinese and Japanese. |
Where does the data go?Why it mattersYour Council will ask this first. | Data is stored in Australia. Conversations are processed by the AI provider in the United States. We will walk your team through the detail. |
Who stands behind it?Why it mattersYou need a partner who answers the phone. | careplans ai, a founder-led Australian company. Ethics approval completed. Signatory to the Rome Call for AI Ethics. A completed trial with Catholic Healthcare. Accepted into the Stone and Chalk Buy Australian AI program. |
How can we test it safely?Why it mattersSmall, defined steps respect your governance. | A short pilot, set out below. |
A short, defined pilot with a small group of home care clients. We agree the outcomes with you before the first call and size it to suit the review process of your AI Clinical Council.
On a 20 minute call we agree four things: which service starts, how many clients, which outcomes your Council wants to see, and what evidence it needs from us first.
Booking page: https://careplans.io/book/andrew
Andrew Payne, Founder and CEO, careplans ai
Email: andrew@careplans.io
Website: https://careplans.ai