A private analyst, in your building, that cites the page.
For the documents you can never put in the cloud. RAGent is a private AI appliance that reads your confidential documents — all of them, going back decades — and answers questions in plain English, showing the exact page each fact came from. Nothing ever leaves your network.
Book a free scoping visitBuilt for law firms, investigators, accountants, and any regulated team whose documents cannot leave the building.
The problem
Your team sits on thousands of pages of sensitive material — case files, contracts, ledgers, client records — and needs to search it, summarise it, and pull out answers fast. Public AI tools are tempting, but you can't use them here:
Confidential by law
The documents are privileged, evidential, or protected by data-protection rules. Sending them to a third-party cloud AI isn't a preference call — it's a breach.
Too big for a chatbot
You can't paste an archive into a chat window, and a chatbot has never read your files. The knowledge that matters is in your building, not on the internet.
No proof
A confident answer with no source is useless for court, disclosure, or audit. Every answer needs to point at the page it came from.
How it works
One machine on your network. Everything happens inside your building — the dashed line is the whole product.
Your documents
Document system, shared drives, email archives, scanned paper. They stay exactly where they are.
The RAGent appliance
One machine on your network. It holds a searchable index of every document, and the AI models that read and answer.
Your staff
Any browser on your network. No software to install. Ask a question, get a cited answer.
- A member of staff types a question.
- The system searches the index and pulls the relevant pages.
- The AI writes an answer using only those pages.
- The answer arrives with citations — click one, see the source page.
☁ The cloud: not involved. The appliance works with the internet cable unplugged.
The grounding rule — the AI is only allowed to answer from the pages it just retrieved. If the answer isn't in your documents, it says "not found in your documents" rather than making something up. That, plus the citation on every claim, is the difference between a clever toy and a professional tool.
How is this different?
Cloud AI tools are good at what they do. They just can't do this.
| ChatGPT / Claude | Microsoft Copilot | RAGent | |
|---|---|---|---|
| Where your documents go | Their cloud (US) | Microsoft's cloud | Nowhere. They stay in your building. |
| Has it read your archive? | No | Partially — what lives in Microsoft 365 | Yes — all of it, including scanned paper |
| Cites the exact page? | No | Links to files, loosely | Yes — every claim, click through to the page |
| How it's priced | Per user, monthly, forever | Per user, monthly, forever | Buy the machine once. Whole firm uses it. Fixed monthly support. |
| Works offline / air-gapped | No | No | Yes |
| Best at | General knowledge, drafting from scratch | Email and Office documents | Your confidential archive, with proof |
We'll say it plainly: if you want help writing marketing emails, use Copilot — it's good at that and we don't compete with it. RAGent is for the documents that can never go where Copilot lives.
What it won't do
Straight answers up front, because that's how we work in the room too.
It won't replace your legal research tools
It answers from your documents — matters, contracts, correspondence, policies. It doesn't know last month's case law. Westlaw tells you what the law says; RAGent tells you what your firm knows.
It won't ask you to take its word
No AI is perfect. That's why every claim carries a citation you can click and check — you never have to trust the machine, only the page it shows you.
It won't lock you in
You own the machine and everything on it. The software stack is open source — no licences to expire, no cloud service that can be switched off. If we vanished tomorrow, it keeps working.
Products
Three made-to-order private AI appliances. Same software, same guarantees — sized to how many people ask questions at once, and how deep the machine needs to think.
Desk
5–20 staff · one team
A small practice or a single team. A quiet tower under a desk that answers a few people at a time.
Practice
20–100 staff · firm-wide
Whole-office use for a mid-size firm. Deeper reasoning, bigger archive, more people asking at once.
Vault
100+ staff · heavy caseloads
Large archives and deep research all day — litigation-scale document sets, whole-building use.
Pricing is fixed and in writing, and it starts with a free half-day scoping visit: we sit with you, review your document workflows, size the right machine, and you get one quote covering hardware, setup, training, and monthly support. No per-user fees — you bought the machine; five users or five hundred, no licence meter running.
From first visit to first answer
It's an appliance, not an IT programme. No servers to buy, no cloud accounts, no six-month project.
Scoping visit half a day · free
We sit with you, run through a structured requirements review, and size the right machine.
Fixed quote within days
One document: hardware, setup, training, monthly support. In writing.
Build & test 2–3 weeks
We assemble your machine, install the software stack, and burn-test it before it ever sees your building.
Install & index on site
The appliance needs a power socket and a network port — no screen, no keyboard. We connect it, point it at your document stores, and it indexes. Large archives index over a few days, unattended.
Training & handover
We train your staff in the room, hands-on, with your real documents. Then it's yours.
Book a free scoping visit
Half a day, at your offices, no obligation. We review your document workflows, show you the system answering on real documents, and you get a fixed quote in writing.