Claude and ChatGPT on business accounts, with your documents and your rules.
Your team is probably using AI already, on personal accounts. We move it onto business plans that don't train on your data, load your documents and templates, and write down what goes in and what never does.
Enquire nowFirms that have seen what these tools can do and want the benefit with the risk under control.
Usually one of these has just happened
- You found out someone pastes client emails into a personal ChatGPT.
- A client asked how you use AI on their work.
- You want everyone using the same tool in the same way.
What directors tell us
“Everyone uses it differently, and I can't see what goes in.”
“It writes well, but it doesn't know our clients or our templates.”
“I don't know which plan to buy, or whether it's safe.”
What changes
Answers in your firm's style
Letters, client updates and reports come out in your format, because your templates are loaded.
Client data on the right contract
Business plans keep your data out of model training. Personal accounts don't promise that.
One way of working
Shared projects and a short policy mean a new starter uses it the way everyone else does.
One workflow, for your kind of firm
Law firms
Draft first-pass letters and summaries from your own precedents.
Measured as minutes per first draft
Accounting and audit
Answer routine client emails from your own guidance notes.
Measured as time to first reply
Engineering consultancy
Summarise standards and long documents with page references.
Measured as hours of reading per report
We aim to automate the repeatable part of a job, often 60 to 80% of it, and leave the judgement to your people on purpose.
How it runs
01 · Week 1
Choose
We test Claude and ChatGPT on your real tasks and recommend a plan per role, with the cost per seat in writing.
02 · Week 2
Set up
Accounts, admin controls, data settings, your documents and templates, and shared projects for each team.
03 · Week 2 to 3
Train
Short sessions per team on their own work, a one-page policy and a record of who was trained.
How we scope it
- The result
- How many people use it each week, and how long the agreed tasks take, measured in week 1 and week 6.
- What we need from you
- Your templates and precedents, a list of what must never go in, and one champion per team.
- What done looks like
- Every team has saved prompts for its three most common tasks, and the account settings are checked.
- What a person decides
- Everything that goes to a client is read and signed off by a person.
- The price
- Fixed and written down before we start. We put a number on what the job costs you now, then price the work so you can see the payback before you say yes.
Put a number on it first
Before we price anything, we work out what the job costs you today: hours a week, times what those hours cost, times working weeks.
Example
Two people spend six hours a week chasing documents. At €45 an hour over 48 weeks, that is about €26,000 a year.
| Hours a week | 12 |
|---|---|
| Cost per hour | €45 |
| Working weeks | 48 |
| A year | €25,920 |
How we do it
- Tested on your work first
- We pick the tool by running both on the jobs your team actually does, then show you the results.
- Settings checked
- Training on your data stays off, admin controls are on, and data residency is set where the plan offers it.
- A policy people use
- One page, plain English, written with your team. Which tool, which account, and what never goes in.
Worth knowing
Microsoft and LinkedIn found that 78% of people who use AI at work bring their own tools. At small and mid-sized companies it was 80%.
Microsoft and LinkedIn, 2024 Work Trend Index · Checked 2026-11-12
Anthropic says it doesn't use inputs or outputs from its commercial products, such as Claude Team and Enterprise, to train its models by default. OpenAI says the same for ChatGPT's business plans.
Anthropic Privacy Center · Checked 2026-11-12
Questions
Enquire nowClaude or ChatGPT?
It depends on the work. We test both on your tasks and recommend one, or each for different teams.
What does it cost?
The vendor's price per seat, paid to them, plus a fixed fee for our set-up and training.
Is client data safe in it?
On a business plan with the settings we check, your data isn't used to train models. For data that must never leave your control, there's private AI.
How do you price it?
A fixed price, written down before we start. We first put a number on what the job costs you now, so you can see the payback.
