For UK charities · Case notes & support records
Summarise a case just by asking
Months of notes into one clear summary. And nothing leaves the building.
I build UK charities private AI assistants that answer questions from the systems they already run, without any of it leaving their control. Each one is built for one job; this one reads the case notes in your case management system.
You ask in plain English, it gives you a structured summary, and every point links back to the exact note it came from.
The quick way to do this today is to paste the record into ChatGPT, which sends a named, often at-risk person's most sensitive details, their health, their safeguarding history, their family circumstances, to a US company, into logs you can't see, with no lawful basis to be there. And a public chatbot can quietly drop a risk or invent a detail, which in a case record is a safety failure, not a typo.
The assistant does the same job without either happening.
There's also one that reads across all your systems at once. Below is this assistant doing its job: getting a colleague up to speed on someone you support, without the record leaving your systems.
What you'd ask it
These run against a made-up young person's welfare notes, held in the Fenmere Trust's case management system (CiviCase), so I can show you without a real person on screen. Each one starts with what you're actually trying to do, then the question you'd type, then the assistant answering and citing the notes it drew from.
A colleague is off sick and you're covering one of their young people today, with months of notes you've never read.
You'd open the assistant and ask:
"Summarise Katie Brennan's case so I can pick up where a colleague left off."
You've been on leave for a few weeks and want to know what's actually changed for a young person, rather than re-reading everything from the start.
You'd open the assistant and ask:
"What's changed for Katie Brennan in the last month?"
Before you hand a young person on to a colleague you want to be sure nothing important is still hanging, and you want to see exactly where each point came from.
You'd open the assistant and ask:
"Pull out any risks or actions still outstanding from Katie Brennan's notes."
These demos read case notes held in CiviCase (CiviCRM's case management), which is just what I used for the demo. Your notes will probably live somewhere else, and that's fine. Lamplight, Charitylog, Views, a shared drive: if it has a way in for the assistant to read, and almost all of them do, the same summaries work the same way.
Want an assistant like this for your own case management system, or have a question of your own in mind? Tell me what your team needs to get up to speed on →
Where the record goes, and why you can trust the summary
Nowhere it shouldn't. The assistant runs in an account you own. The notes, your questions and its summaries are never sent to ChatGPT, Gemini, Claude or any other outside service. It reads the record with its own read only access, so it can summarise but cannot change, delete or export anything. If it tried, your systems would refuse.
And it doesn't work from memory. It reads the actual notes each time and every point in the summary links back to the exact note it came from, so a worker can check it in seconds and stays the one who decides. It draws only from what the notes say, rather than smoothing a risk away or inventing a tidier story, and before any of it reaches your team it is tested against records where the right summary is already known.
On a case record, that faithfulness is the whole point. Here's how I test AI assistants and what the testing catches →
Common questions
Can I use ChatGPT to summarise case notes?
You can paste text in, but you shouldn't with real client records. Case notes carry named individuals' most sensitive details, and the public version of ChatGPT sends them to OpenAI in the United States, keeps them in logs you can't see, and may use them to train future models. For a UK charity that's personal data leaving your control with no lawful basis to justify it.
Is it safe to put client records into ChatGPT?
No. Records about service users, especially involving health, mental health, abuse, children or immigration status, are personal and often special category data. Pasting them into a public tool sends them outside your charity entirely. You'd need a lawful basis, a contract with the provider and usually a data protection impact assessment; a paste into a public tool has none of those.
Can AI summarise case notes without breaking confidentiality?
Yes, if the summarising happens inside your own systems rather than a public tool. A private build reads the record where it already lives and produces the summary there, so the confidential information is never sent to a third party in the first place. That's the difference between a tool that helps and one that creates a breach.
Can the AI miss or change something important in a summary?
That's the real risk with a public chatbot, which can drop a detail or invent one. A private build is grounded in the actual notes and cites the source for every point, so the worker can check in seconds that nothing was softened or added. It draws only from what the notes say rather than producing a plausible-sounding story.
Does this replace the worker's professional judgement?
No, and it shouldn't. The assistant does the reading and produces a summary you can verify against the source notes; the worker stays the decision-maker and remains accountable. It's there to save the hours spent re-reading a record, not to make the call for you.
Can the AI edit or delete our case records?
No. The build reads a copy of the record, so it can summarise but cannot write, delete or export anything. It also shows the note behind every line of the summary, so anything it tells you can be checked against the original.
Got a question that isn't here? Ask me directly →
One job is just the start
Everything on this page is one assistant, built for a single job. That's deliberate: most charities prove it on the one task that's hurting, then add the next.
The most capable assistant I build reads across your CRM, your finance system, your spreadsheets and your documents at once, and answers questions that cut across all of them. One question, one plain English answer, every figure linking back to the real record it came from.
See what that looks like when it all joins up. Watch the full walkthrough →
