AI consultancy for owner-led businesses
The AI running our business can run yours.
It answers our phones, routes our enquiries, runs our customer portals and takes our payments. Everything we build for you, we run ourselves first.
We help owners and managing directors work out where AI genuinely earns its keep, then build it into how the business runs. It starts with a fixed-price review.
Where it usually starts
If you could add ten good people to your team tomorrow, what would they work on?
AI now makes much of that capacity real, but plenty of what's sold to businesses doesn't pay back. These are the jobs where it usually does.
The same customer questions, answered by hand all day.
An assistant that answers from your own documents and passes the rest to a person.
Leads that arrive at 9pm and get a reply at 10am.
Enquiries answered and routed the moment they land, day or night.
A weekly report stitched together from several systems.
A report that assembles itself from the systems you already use.
Quotes that take hours to prepare.
Quotes drafted from your price list and past jobs, for a person to check and send.
Information that lives in one person's head.
A searchable record of how the business works, that the whole team can ask.
What we do
Advise, build, then keep it running.
01 Advise
Start with a fixed‑price review.
Two weeks and a written plan: the jobs worth handing to AI, what each costs and the payback, and what to leave alone. No obligation to build.
A fixed price, agreed before we start. No obligation to build.
02 Build
Ship AI that does a real job, in your systems.
Products, integrations and automations, built with AI agents and reviewed by an experienced engineer, in your repositories from day one.
Hours from our public cost model, at a rate printed on the quote.
03 Operate
Keep it working as models, prices and your business change.
Monitoring, fixes and improvements after launch, so the system you rely on keeps doing its job.
A monthly arrangement agreed in writing; usage passed through at cost.
Case study
A homeowner asks for a price. Nobody has to be at a desk.
A marketplace we built and run for home-heating fuel in the US. Requests arrive at any hour and reach the local dealers who can serve them, and the dealers take it from there in their own portal. When a call would help, the phone does the first part. It is the kind of system we build for clients: enquiries handled when they land, with a person where a person matters.
- requestA homeowner asks for a delivery price, at any hour.✓ stored with their consent
- matchMatched to the dealers that cover their ZIP code.✓ automatic
- portalEach dealer sees it in their own portal and claims it.✓ nobody approves it by hand
- phoneWhere a call helps, our phone rings, reads the request out, and one key connects us to the dealer.✓ recorded and transcribed
- nowlive
Who you talk to
Oliver Wakefield-Smith has spent more than twenty years building software for businesses, independently for most of that time. He now directs a team of AI agents and reviews what they ship.
You deal with him from the first call to the last release. He writes the specifications, directs the agents and reviews everything they ship, and he answers for the result. Clients include Morrison Energy Group.
Tell us what is eating your team's week. Thirty minutes is usually enough to know whether AI is the answer.