I build AI that does real work in real industries — and help others do the same, whatever your starting point.
AI is not able to do quality work straight out of the box, no matter how strong it is. It is not always a game of raw intelligence — it is a game of equipment.
Agents do not fail because they are stupid. They fail because you do not give them the right tools, equipment, and access to knowledge. Get those three right and the model becomes a colleague.
I have taken this approach into the UK construction industry and built Olive: an autonomous assistant that measures, prices, drafts, and reports — accurately, end to end.
It works because of the toolbelt around it.
If you are technical, here is what AI means for your stack: AI shouldn't be slotted into your existing Python and JSON infrastructure. Flip the relationship.
The AI should be the thing doing the thinking. Your Python mechanics and JSON schemas should exist to complement the AI — used as guidelines and playbooks, not as dependencies the model is forced to obey.
That inversion is the unlock. It lets your tech stack think and work on problems like a member of the team — instead of sitting around waiting for a colleague to start repetitively typing on a computer.
And if you're the kind of technical person who points to hallucinations and AI errors as a reason not to adopt — I've heard it a lot — you may as well stop hiring humans, because they make mistakes too.
The right answer is not to refuse adoption. It is to be pro-adoption and pro-risk-mitigation at the same time: train the model to be brilliant, and make the sandbox around it even better. A great employee in a well-designed environment will outperform a perfect one with no guardrails — and there is no perfect one.
For the last few decades, the only way to use the tools on your computer was to type and click. Now it will be speaking to your computer like you speak to a colleague.
I offer consultancy on the steps to take so your team gets an assistant that elevates the work they already do. Advice is available at any experience level.
The common thread: keeping up with AI without being left behind by it.
I built Olive on a set of concepts: design tools for AI users — not human users; treat Python and JSON schemas as playbooks rather than dependencies; and put the model in the driver's seat with a great sandbox around it.
Those concepts work for any business in any industry, and I have a passion for seeing them out in the wild — not just in construction.
If you want to talk about what that could look like for your team, book an hour with me.
Book a 1-hour consultation →Or write to me directly: [email protected]