AI transformation
From scattered ChatGPT to agentic-first
You’ve seen what AI can do in flashes. A good draft here, a task that took ten minutes instead of an hour there. What you haven’t got is a business where that’s the norm rather than the exception. This is the engagement that builds that, end to end, with your team able to run it afterwards.
Who this is for
Founder-led service businesses that have outgrown ad-hoc AI.
Typically 10 to 200 office staff. Profitable. Run by the person who built it. AI is being used, but it’s a scatter of personal habits and one-off experiments, not something the business runs on.
If you’re still working out whether AI is worth taking seriously, this probably isn’t the right time. If you’ve decided it is, and you want it built properly rather than bolted on, that’s exactly what this is.
The businesses that win with AI aren’t the ones with the most tools. They’re the ones that built it in the right order.
Why the order matters
The gap between the leaders and everyone else is already this wide.
A small group of companies got their foundations right early, and the returns have been compounding on that head start ever since. Most small and mid-market businesses have done little beyond handing round a ChatGPT licence. The few that ran a proper pilot mostly found that piloting AI and getting it to pay off are different things. Either way, almost nothing has changed for them, and that’s the gap.
the productivity AI leaders get from their workflows, around 25 to 40%. The median large company that's invested gets 3 to 7%. Most everyone else, nothing.
PwC 2026 AI performance study
of company AI pilots deliver little to no measurable impact on the P&L. Most of the spend never shows up in the numbers.
MIT's State of AI in Business study, 2025
How the engagement runs
Audit, build, train. In that order.
Audit
Half a day inside your operations
Before anything gets built, I map where your business actually bleeds time, where information sits in silos, and where decisions move slowly. You get a prioritised roadmap within 24 hours, with the quick wins at the front and the bigger builds sequenced behind them.
Half a day covers most businesses. If yours is larger or more tangled it can take a little longer, and I’ll tell you that before we start. Nobody’s pretending a 200-person operation gets solved in an afternoon.
Build
Foundations, then workflows, then agents
We do it in order. Clean, connected foundations first. Then the repetitive workflows get automated, which is where the time starts coming back. Then agentic systems on top of that, doing the work that used to need a hire. Skipping the order is why most implementations fall over, so we don’t.
Train
Your team learns to run it
As the system goes in, your people learn how it works and how to keep improving it. The point is a business that owns this, not one that depends on me to keep the lights on. If team training is the only thing you need right now, I can do that on its own as a starting point.
Optional
I stay on, if you want me to
Some businesses want a hand once the system is live, for maintenance and for the next round of builds. There’s an option for that. Plenty of businesses don’t need it, and that’s a fine outcome too.
The methodology behind it
Every engagement runs on the same framework.
Six stages, from audit through to an AI-first culture. The framework is the same for every business. How it gets applied is always specific to yours.
I built and ran Oh Crap for nine years and sold 58 million bags before exiting. I know what it’s like to run the thing, not just advise on it. That’s the seat I’m building from.
Henry Reith
Book your AI-First Assessment
A 30-minute call to map where your business sits on the AI maturity ladder and whether working together makes sense.
1 advisory spot left this month.
- Designed for 10 to 200 staff service businesses
- Workflows, not prompts
- Practical implementation, not workshops
- No random tool stack, just what your business needs