Work with us
AI that earns its place, in three safe steps.
Most AI projects start with a tool and hope it fits. We start by proving the return, with evidence you and your board can decide on, before you spend on anything else.
Everything in the hub stays free. These are paid services, and they are how we pay for it.How this hub is funded
- Writing up visit notes£5,070
- Monthly returns to commissioners£2,592
- Staff questions about policies£1,440
Yearly saving, top three£9,102
An illustration, not a real client.
The three steps
Evidence first, fixes second, then the solution.
Each step stands on its own, and you can stop after any of them.
- Start here
Step 1: Prove the return first
Know what AI is worth to you, and what stands in the way, before you spend anything.
Paid service
Step 2: Fix what is blocking you
The gaps step 1 found, fixed, so you are covered with staff, families and regulators.
Paid service
Step 3: Put the right solution in place
More time for care, from whatever gives the best return, even when that is not us.
Paid service
Step 1 of 3
Prove the return first
Know what AI is worth to you, and what stands in the way, before you spend anything.
Start herePaid service
What you get
- Where your team’s time really goes, task by task
- What could realistically be automated, and what could not
- The yearly saving for each option, ranked
- What is blocking you: policy, data protection, suppliers and data
- A plain report your board can decide on
What we do
We map how the work is really done today: each task, who does it and how long it takes. Then we work out what could realistically be automated, what that would save, and what is blocking you from doing it safely: missing policies, data protection gaps, supplier agreements and data that is not ready.
How it works for you
We set it up
We agree with you which teams and tasks to look at, and set up a simple time log.
Your staff log their time
For an agreed period, staff note what they spend their time on. We support them throughout and do as much of the work for them as we can.
We do the analysis
We turn the logs into hours and cost for each task, work out where AI could realistically help and where it could not, and list what is blocking you.
You get the ranked results
A plain report of the options, the yearly saving for each and the blockers to fix, to take to your board.
What the results look like
- Hours a year it could free up
- 659
- Yearly saving across the options
- £9,912
- Task kept with people, on purpose
- 1
Example totals
| Task | Hours a year now | Hours it could save | Yearly saving | Option |
|---|---|---|---|---|
| Writing up visit notes | 1,300 | 390 | £5,070 | Carers speak their notes and AI drafts the record. The carer checks it before it is saved. |
| Monthly returns to commissioners | 240 | 144 | £2,592 | AI fills a template from your records, and a manager checks every figure. |
| Staff questions about policies | 200 | 80 | £1,440 | A search tool over your own policies that shows where each answer comes from. |
| Finding rota cover | 300 | 45 | £810 | Low priority. A small saving, and a coordinator still has to check every suggestion. |
| Reviewing care plans | 400 | None | None | Keep with people. Decisions about someone’s care stay with staff who know them. |
An illustration, not a real client, and not a promise of what you will save. Hours are valued at the cost of the staff who do each task. Your own report uses your own figures.
Want to measure a task yourself first? Our free time tracker and guide show you how.Admin time trackerMeasuring a task before you change it
Step 2 of 3
Fix what is blocking you
The gaps step 1 found, fixed, so you are covered with staff, families and regulators.
Paid service
Step 1 shows what is blocking you. Step 2 is only about fixing those blockers. We fix what the report found, fitted to your services and checked with you before anything goes live.
An AI use policy
Rules your staff can follow, fitted to your services.
A data protection impact assessment
For any use of AI that involves personal information.
Data processing agreements with suppliers
Checked, so you know where data goes and who is responsible for it.
Data and documents ready for AI
Policies, templates and records tidied, and anonymised where needed, so a tool can work from them well.
You can start any of these yourself with our free templates. In step 2 we fix the ones step 1 found with you, fit them to your services and check that they hold up.
Step 3 of 3
Put the right solution in place
More time for care, from whatever gives the best return, even when that is not us.
Paid service
Not always a build. We recommend the route that pays back best for your problem, and say so plainly when that is not something we make.
One of three routes
A tool we build for you
Built around your problem, and tested with your staff on realistic work before anyone relies on it.
Setting up what you already have
Many organisations already pay for tools, such as Microsoft 365, that can do more than they are used for. We set them up properly and safely.
Choosing a product you can buy
If something on the market does the job, we help you choose it, check the supplier and set it up.
Whichever route you take
Training and support for your team
So staff know how to use it, how to check its work and when not to use it.
We also make Eva, a voice check-in service for people between visits. If it fits your problem we will say so, and if it does not, we will not suggest it.About Eva
Why start with step 1
The difference evidence makes.
Going straight to a build
- Built around a guess about the problem
- Can add work instead of saving it
- No proof of return, and a better option missed
- Policy and data protection gaps found too late
Starting with step 1
- Decisions backed by evidence of the return
- The saving for each option known up front
- Blockers found early, not after go-live
- Your team stays in control: AI drafts, people decide
Questions
What people ask before step 1
How much time does step 1 take from our staff?
Staff note their tasks as they go, in a simple log we set up. Before it starts we agree with you which teams take part and for how long. We do the rest: setting it up, answering questions, reminders and all of the analysis.
Does the time log involve personal information?
It should not. The log records tasks and how long they take, not details about the people you support, and we set it up that way.
What if the answer is that AI will not help?
Then that is what the report says. Knowing where AI will not pay back is as useful as knowing where it will, and you find out for the price of step 1 rather than a project that does not work.
Do we have to go on to steps 2 and 3?
No. Each step stands on its own, and you can stop after any of them.
A supplier is already pitching us a tool. Can you still help?
Yes. Step 1 shows whether the tool is aimed at the task that would save you most, and what policy and data protection work it would need. Step 2 fixes that. We will tell you plainly if it is the right choice.
Who makes the decisions once AI is in place?
Your staff. Whatever we put in place, AI drafts and sorts, and people check and decide. Decisions about someone’s care stay with the people who know them.
How is this different from the cohort programme?
The steps are done for your organisation alone, on your own tasks and in more depth. The cohort is a group programme: 8 to 12 organisations learn about AI together and shape one shared tool. It suits you if you would rather learn alongside others.
Prefer to learn with others?
The cohort programme
Over three months, 8 to 12 care organisations learn about AI in care together through webinars and Q&A sessions, agree on one shared problem, and shape an AI tool that we build for it. Each organisation keeps using the tool when the cohort ends.
The price is to be announced. Dates for the next cohort are to be announced.
Start with the evidence.
Tell us a little about your organisation and we will arrange step 1, or say plainly if something free in the hub would serve you better.