AI consultancy that has to pay its way, or we tell you not to buy it.
We find where AI would pay in a UK SME (or a team inside a bigger business that runs like one), then deliver it through three services: AI strategy, AI development (assistants and chatbots grounded in your own data) and AI automation. We also write down what AI should be kept away from. We ship AI in our own products, so the advice comes from running it, not reading about it.
A costed shortlist of AI uses that pay, not a slide deck
Assistants grounded in your data, running costs known up front
A written do-not-touch list that saves expensive experiments
How we can help
AI Strategy & Readiness
AI strategy and readiness reviews: where AI pays in your business, and where it doesn't.from £2,500
AI Development
AI development: assistants and chatbots grounded in your data, not the internet's guesses.from £6,000
AI Automation
AI automation: the hours lost between your systems, given back.from £3,500
The hype filter.
Five claims we hear in AI sales meetings, and the verdicts we actually give. Press one; we will be straight with you.
Our verdict on “AI will replace your whole team”
No. It removes the tedious slices of jobs, not the jobs. The firms getting real value are freeing good people for better work, not planning redundancies around a demo.
This is the same filter we apply on a scoping call, which is why some conversations end with us advising you not to buy anything.
Where AI pays back first.
Four places the hours usually leak. Pick the one that sounds like your business and read the shape of the first project we would propose there. A shape, not a promise: it only becomes numbers on a scoping call.
- 01 · The goal
Invoices, forms and PDFs stop being re-keyed into your systems by hand.
- 02 · The first slice
One document type first, extracted straight into the system it belongs in, tested against real historical cases before it touches live work.
- 03 · What proves it worked
Hours a week measurably returned, with every automated action logged and reversible so finance can see exactly what it did and why.
Every one of these starts small and fixed-price, because proving value on a thin slice beats betting the year on a big build.
AI Consultancy, answered
We keep being told to 'do AI'. Where do we actually start?+
Not with a model. We start with the work: where your team spends hours on things a machine does well, where decisions wait on information that exists but is buried, and where customers wait on answers you already hold. That audit usually surfaces two or three uses that pay for themselves quickly, and a longer list of things AI should be kept away from. You get both.
Will our data end up training someone else's model?+
No. We build on APIs and configurations where your data is not retained or used for training, and we put it in writing. For sensitive workloads we design so the sensitive part never leaves your systems at all.
Do you build chatbots?+
When a chatbot is the right shape, yes, grounded in your real documents and data so it answers with your facts rather than the internet's guesses. Often the better win is less glamorous: automation that removes the copy-paste work between systems, or drafting that a human approves. We recommend the shape that pays, not the one that demos well.
What does an AI project cost?+
A readiness review is a fixed, low-four-figures engagement. Builds are scoped like any software project, and because most AI features ride on an existing app or site, they are often smaller than people expect. Ongoing model costs are estimated up front so there are no surprise bills.
Find out where AI would pay in your business, before you spend on it.
- Describe the problem, not the technology. A paragraph is enough. You'll hear back within one working day.
- One scoping call. With the consultant who would do the work. Expect to be asked what the problem costs you today.
- A fixed figure in writing. Or a plain 'this isn't an AI problem', if that is what the call finds.