A fractional AI lead. Someone senior who's actually built it.
Most firms this size have nobody senior who understands AI from the inside, so decisions get made by whoever read the most LinkedIn posts, or by the vendor with the best lunch. This is a day or two a month of someone who has shipped it at industrial scale and run two businesses on it, sitting on your side of the table.
What it looks like
- In the room. At the leadership meeting when AI comes up, as your adviser, not a supplier. Asking the vendor the questions they don't want asked.
- A roadmap you own. Six to twelve months, prioritised by what pays back, with real costs and real savings, reviewed monthly. Written so your board can read it.
- Saying no. To the £2,000-a-month platform that's a wrapper, to the pilot that can never go live, to the tool that would put your client data somewhere it shouldn't be. Half the value is the nays.
- Governance that fits. Who can use what, what goes where, what needs approval, what the audit trail looks like, and a one-page policy your staff will actually read. UK GDPR as the floor, not the ceiling.
- Hands on when it counts. Vendor selection, data readiness, supplier oversight, and building the thing ourselves when that's the honest answer. Including bespoke MCP work where you've got your own software.
Who it's for
- Small manufacturers and engineering firms with machines throwing off data nobody looks at.
- Professional services: accountants, solicitors, surveyors, agencies. Document-heavy, process-heavy, nervous about client data.
- Multi-site businesses where the same admin happens five times.
- Anyone who's been quoted five figures for "AI transformation" and would like a second opinion before signing.
Why this person
Thirty years as a software architect. The biggest AI product shipped is an assistant inside a calibration platform used by major car makers' engineering teams, built from the model up. Two businesses founded or run on the same methods being recommended, so the advice has had to pay for itself. And the same "aye or nay" habit as the rest of this site: if something isn't worth doing, that's what you'll hear.
How it starts
With the same free look-round as everyone else, scaled up: a half-day with the leadership team instead of an hour with the owner. You get a written view of where you are, the three things worth doing first, and a proposed shape for the engagement. If it doesn't stack up, you've had a useful morning.