AI consulting
AI consulting that starts with your business, not the model
Most AI advice arrives with a product attached to it. Mine does not, because I do not resell anything. I earn money building the things that turn out to be worth building, which means I have no reason to talk you into the ones that are not.
The work usually starts with your processes rather than the technology: where time actually goes, which of those steps a model could genuinely take over, and which ones only look like AI problems until you examine them.
Get in touchWho this is for
- Someone has told you AI could help, but not where or how much.
- You have a shortlist of ideas and want them sanity-checked before you commit budget.
- A pilot has stalled somewhere between an impressive demo and anything you could put in front of customers.
- You want a straight answer on whether a project is worth doing at all.
What you get
A ranked shortlist
Your candidate use cases, ordered by what a working version would save against what it would cost to build and — more often forgotten — to run every month afterwards.
A technical opinion
Which Azure AI services actually fit, what shape your data needs to be in first, and where the integration work with your existing systems is going to hurt.
A roadmap you can budget
A sequence that starts with whichever piece proves or kills the idea fastest, so you find out cheaply rather than at the end.
An honest no where it applies
A good number of problems are better solved by a database query, a rule or a better form. When that is the case here, I will say so.
Related work
Common questions
What does an AI project cost?
It depends far too much on scope for a number here to mean anything, and anyone quoting one before understanding your case is guessing. What I can tell you is that the first conversation is free, and that I would rather scope a small first piece you can evaluate than a large one you have to trust.
Do we need our data in order before we start?
Less often than people fear. Some use cases genuinely need clean, structured data first; others work fine against documents and systems exactly as they are today. Working out which of the two you are looking at is part of the first conversation.
Can you work with our existing Microsoft setup?
That is the usual starting point. Most of my work sits on Azure, .NET and Dynamics 365, so an existing Microsoft estate is an advantage rather than something to work around.