AI audit & scoping

The problem

Everyone tells you that you need to “do AI”. Nobody tells you where to start, what it actually costs, or which projects will fail because your data isn’t ready.

The result: either nothing ships, or three pilots run in parallel and all three end up in a drawer.

What I do

I work with you part-time for two to three weeks:

1. Business interviews — five to eight conversations with the people doing the work today. That is where the real use cases surface, not in the steering committee.
2. Technical review — the state of your data, your infrastructure, and your regulatory constraints (GDPR, AI Act, regulated sectors).
3. Prioritisation — every use case is costed: expected value, effort, risk, dependencies. Projects that fail the filter are ruled out explicitly, with the reason.
4. Readout — a two-hour session with your decision makers, then a document you can use without me.

What you get

• A short list of two or three use cases to launch, in order.
• A realistic budget and timeline for each.
• The technical prerequisites, and what is blocking today.
• A reasoned position on what not to do. That part usually saves the most money.

What it is not

It is not an 80-page consultancy report. It is a 15-to-20-page document that a CTO or a founder reads in an hour and can decide on.