For UK charities · Internal knowledge & documents
Yes, and it might be the easiest win on this list. The catch is where the documents go.
Your charity already holds the answers to most of what your team asks all day. They're just scattered across drives, inboxes and the heads of people who may not be here next year. The job is to make all of it findable, without handing it to a stranger to do so.
Here's why your charity already has the answers but can't find them, and how a private assistant that's read everything gives them back, with none of it leaving.
It usually starts as something small. Someone needs to know the sign-off limit for spending, what a funder's grant agreement requires, or what the board actually decided about reserves. The answer exists in a document somewhere, but finding it means hunting through folders or interrupting the one person who knows. So increasingly people take a shortcut: they paste the document into a public AI tool like ChatGPT, Gemini or Claude and ask it. Nothing about the intent is reckless. The trouble is on both ends: the knowledge is hard to reach, and the fast way to reach it leaks it.
Charities run on a few people who simply know how things work. When one of them leaves, and turnover in the sector is high, the knowledge leaves with them. New staff and trustees then spend months piecing together what was once obvious, and some of it is never recovered at all.
Documents live across Google Drive, SharePoint, shared inboxes and the odd laptop, in versions no one is sure are current. Decisions end up made on out-of-date information or a best guess, because tracking down the right document took longer than the decision was worth.
To get a fast answer, people paste in the real document: board minutes, HR files, management accounts, legal advice, funder contracts. Plenty of that is confidential or special category data, and the moment it's pasted in it lands on a US company's servers with no lawful basis for sending it there.
If a new trustee asked a basic question about how your charity works, could anyone point them to the answer, or only to the person who happens to know?
None of this means replacing the knowledge your team carries. It means giving everyone access to it. The answers your charity already holds are genuinely valuable; the job is to make them reachable in seconds without sending anything outside your walls.
Imagine an assistant that has read your whole body of documents and answers any question about them in plain English, citing exactly where the answer came from, with none of it ever leaving your own systems. Three things change.
The knowledge no longer lives in one person's head. When someone leaves, what they wrote down is still answerable, and a new starter or trustee can get up to speed in days rather than months by simply asking.
Ask in plain English and get the answer drawn from the right document, with a citation to the exact source, instead of hunting folders or guessing. No more acting on a version that turned out to be three years old.
Board papers, HR files, accounts, legal advice and funder contracts are indexed inside your own tenancy. The confidential material becomes searchable to your team without ever being handed to a third party to read or train on.
That's private AI for your documents: every answer your charity already has, findable in plain English, on documents that never leave your control.
The build is a private AI assistant, a chat tool that looks and works like ChatGPT but runs inside your own systems, that has read your charity's own documents, your staff handbook, policies, procedures, guidance, board papers and funder agreements, and keeps them in a private knowledge base inside your own tenancy. You ask a plain-English question; it retrieves the relevant passages and answers, citing which document and section each part came from so you can check it. It answers only from what your documents actually say rather than inventing, and it reads only: it cannot change or delete anything.
Your documents can stay where they already live. Whether they're in Google Drive, SharePoint or a shared folder, we set up a connection that brings them into your private index and keeps it current as they change. The content is copied into your own store, not sent out to a public tool, so the assistant stays up to date while the documents never leave your control.
Same question you'd have pasted an internal document into ChatGPT to answer. None of it leaving the building, and an answer you can trace to the source.
I've built a sample assistant against a fictional charity's documents here, to demonstrate the kinds of thing a private AI build can do for you. The documents are entirely made up. Below are a few everyday situations someone might bring to it, deliberately spanning different kinds of document. Each one starts with what you're actually trying to do, then the question you'd type, then a short recording of the assistant answering and citing the document it drew from.
You've just started, and you want to know where you stand on leave without having to collar a colleague on your first week.
You'd open the assistant and ask:
"How much annual leave do I get, and how do I book it?"
You're about to commit to a supplier and you're not certain who needs to approve the spend, or at what level.
You'd open the assistant and ask:
"What's our process for approving spend over £5,000?"
You're a trustee prepping for a meeting and you want to be sure of what was actually decided last time, without re-reading every set of minutes.
You'd open the assistant and ask:
"What did the board decide about our reserves policy, and when?"
A funder report is due and you want to be sure exactly what this grant asked of you, and by when, rather than re-reading the whole agreement.
You'd open the assistant and ask:
"What reporting does our main grant require, and when is it due?"
Want to see it pointed at your own documents, wherever they live, or have a question of your own in mind? Tell me what your team can never find →
You can paste a document in and ask about it, but you shouldn't with anything internal. The public version of ChatGPT sends whatever you paste to OpenAI in the United States, keeps it in logs you can't see, and may use it to train future models. For a charity, board papers, HR files and funder contracts leaving your control like that has no lawful basis. A private build lets you ask the same questions without anything leaving.
No. Board minutes, management accounts, HR records and legal advice are confidential, and some carry special category data about named individuals. Pasting them into a public tool sends them outside your charity entirely. A private assistant indexes them inside your own systems instead, so they become searchable to your team without ever being handed to a third party.
Yes, and it's one of the most useful starting points. A private build reads your handbook, policies and procedures and answers staff questions in plain English, citing the exact policy and section, so people get the current answer in seconds instead of hunting a shared drive or asking around. Nothing is sent to a public tool to make that work.
Yes. Your documents can stay where they already live: as part of your build we set up a connection from your Google Drive, SharePoint or shared folders into a private knowledge base inside your own systems, and keep it current as documents change. The content is indexed in your own tenancy, not sent to a public tool. This is a connection tailored to where your documents actually are, set up for you, rather than an off-the-shelf button.
It shouldn't, and that's the point of grounding it in your documents. It answers only from what your documents actually say and cites the source for each answer, so you can check it. If the answer genuinely isn't there, a well-built assistant says so rather than inventing a plausible-sounding one.
That's exactly the problem this solves. Today, much of what a charity knows lives in a few people's heads and walks out when they do. Once your documents are in a private knowledge base, what was written down stays answerable for whoever comes next, so a departure or a new starter no longer means months of lost ground.
Got a question that isn't here? Ask me directly →
Tell me who you are and what your organisation does. If any of this sounds like your situation, that's a good place to start. I'll let you know honestly whether I can help. Even a 30 to 45 minute call often leaves people with a clearer picture of the path forward, whether or not we end up working together. From there it's whatever fits: sometimes you don't need me, sometimes a short piece of scoping work makes sense first, and sometimes you already know what you want and we get straight to the build. There's no set process you have to follow.
For context: I work mainly with UK charities and non profits, with chief executives, operations and finance directors, programme leads, and the people who look after data and IT. Respectfully, I don't work with recruitment or development agencies.
Email: peter@peterbrady.co.uk
PNB Technologies Limited
55 Yew Tree Road, Ormskirk
Lancashire, L39 1NT, United Kingdom
Company Reg: 07166600
VAT Reg: GB986726850
© 2026 Peter Brady. All rights reserved.