AI platforms I have built

I do not sell development. It is not on my services list and it is not going on it. This page exists because you are entitled to know whether the person telling you what to build has ever had to build it.

Three platforms for a digital trust and AI company, product definition through to released software. Two are live and one is in build.

The three

• A digital and AI trust platform. Live. Establishing and evidencing trust across a digital and AI estate, so the answer to a customer or a regulator is a system rather than a scramble.
• An AI oversight platform. Live. What is running, who owns it, how it is classified and the evidence behind it. Runtime governance rather than a policy document ageing on a shared drive.
• An AI adoption index. In build. Adoption scored across access, usage, workflow integration, trust and business value. Comparable across business units and repeatable year on year.

Before those, twenty plus data and AI products inside a multi billion dollar business, and enterprise BI platforms carrying twenty five thousand users.

Why this is not a service

Because the moment building is on my price list, every recommendation to build becomes suspect, including the correct ones.

The whole point of the method is to decide honestly whether something should exist. Gate four asks whether to buy, configure or build, and the truthful answer is usually one of the first two. That answer is worth far less if nobody ever checks it, which is why you get the numbers behind it rather than the conclusion on its own.

When I tell you to build something, you get the working: what it costs to build, what it costs to buy, and what it costs to do neither. Take it to somebody else and check. If the answer only holds while I am the one quoting, it was never an answer.

So what happens when the answer is build?

I help you scope it properly, decide who does it, run the selection, and hold them to what they agreed. I will sit on your side of the table while somebody else delivers, and I will read their architecture and tell you what it is going to cost you in three years.

That is the useful role, and it is one a delivery partner structurally cannot play for you.

What building taught me that advising would not

• What things actually cost. Not the estimate, the real number, including the six weeks nobody budgets for after the demo goes well.
• Where specifications leak. A specification handed to strangers is a decision written down and then reinterpreted. Knowing where that happens is how you write one that survives.
• What good looks like in a review. I can read a reference architecture and a delivery plan and tell you which parts are optimism.
• Why most builds fail on adoption, not code. Every platform above was instrumented for usage from the first release, because shipped and used are different achievements.

Where this connects

If you are trying to work out whether the thing in front of you should be bought, configured or built, that is gate four. The Transformation Index takes three minutes and will tell you whether you are even at that gate yet.

If you want the judgement on your side continuously rather than as a one off, that is the fractional and advisory shape.