AI implementation consulting: get it right before you build

What does upfront AI implementation consulting deliver, and which service fits?

AI implementation consulting from Dave Willemstein settles these choices in writing before the build starts: the use case, build or buy, data and privacy, and the change that's needed. The result is a decision document with the decision and the brief for the delivery partner; a concrete question about one application usually fits scoped advice at a fixed price per assignment. Anyone who wants to know first whether the organisation is broadly ready starts with AI Readiness Assessment, which takes 2-3 weeks, for €7,000.

Does AI implementation consulting produce a thick report?

No. AI implementation consulting ends with a decision document, not a thick report: which use case, which build choice, which conditions for data and privacy, what change is needed, and what the delivery partner needs to deliver to succeed. All the frameworks and analysis are then the client's to keep.

Does AI implementation consulting also look at the privacy side?

Yes. AI implementation consulting maps out what data the application needs and also looks at the privacy side. A data protection impact assessment is required under UK GDPR if the application processes personal data with a significant likelihood of high privacy risk, and that question belongs at the start, not just before go-live.

How long does the first conversation about AI implementation consulting take?

Thirty minutes is the length of the introduction call for AI implementation consulting. The client explains where the organisation stands and hears at the end what the adviser would do, even if that means the client doesn't need this adviser.

Source: Dave Willemstein's rates and timelines; primary source: UK GDPR; Information Commissioner's Office: Data protection impact assessments (as of 6 October 2026). Updated: October 2026.

Why upfront advice pays off

An AI project gets expensive the moment you discover an earlier choice was wrong. A use case with no owner, a vendor contract with no exit clause, or a dataset that’s messier than expected: during the build, each of those discoveries costs weeks and budget. Catch the same thing upfront, and it costs a few conversations.

AI implementation consulting isn’t a luxury reserved for large organisations, then. It’s the cheapest phase of the whole project, and the one where you have the most influence over the outcome. For project leads and IT, it means a clear brief for the delivery partner. For leadership, it means a decision that’s backed by evidence rather than resting on a convincing demo alone.

This consulting is one part of the broader guidance through AI implementation, from use case to management. Looking for advice on AI in a broader sense, such as strategy or prioritisation across the whole organisation? See AI consulting.

What I advise on

Four topics decide whether an implementation succeeds. They belong on the table together, because they affect each other: a different use case calls for different data, and different data can flip the build-or-buy choice.

Choosing the use case

The first question isn’t “what can AI do?” but “which problem is costing us too much time, money or quality right now?”. I help assess candidate applications the same way: on value, feasibility, data readiness, change load, and whether there’s an owner in the business who genuinely benefits. The result is a short list, with a reasoned list of what didn’t make the cut. For the kinds of applications that typically deliver results, see which AI use cases deliver results.

Build versus buy

Building it yourself, buying a product, or working with a partner: that choice shapes your costs, your dependency and your agility for years. Build only what sets you apart, buy what’s a commodity, and choose a partner when you want to steer the outcome without owning the technology. Factor in the running costs over several years, and agree upfront how you get out if it doesn’t work. The full decision framework is in build, buy or partner. Choose a partner, and you contract them directly; how I work with that is covered on the partner network page.

Data and privacy

I map out what data the application needs, where it comes from, how reliable it is, and who’s responsible for it. Alongside that, I look at the privacy side. If the application processes personal data with a significant likelihood of high privacy risk, a data protection impact assessment is required under UK GDPR, as set out by the ICO. That question belongs at the start, not just before go-live. I also look at what data you’re allowed to send to an external service, and what agreements with the vendor that requires.

Change management

Every AI application that saves time changes someone’s job. That conversation often gets postponed until after delivery, and then the business case only holds up if something changes that nobody’s bought into. I help name upfront what needs to change, who decides on it, and how you bring along the people who’ll actually use it, with concrete scenarios per role rather than one blanket rollout.

How an advisory engagement works

1. IntroductionThirty minutes: what's going on, and what would I do?
2. ScopingIntake: pin down the question, scope and format in a confirmation of engagement
3. Analysis and adviceUse case, build choice, data and privacy, change
4. Decision documentA reasoned decision plus the brief for the delivery partner

An engagement starts with a thirty-minute conversation. You explain where things stand, I ask questions, and I say at the end what I’d do, even if that means you don’t need me.

After that, an intake scopes the question and sets it down in a confirmation of engagement. A concrete question about one application usually fits scoped advice. Want to know first whether your organisation is broadly ready? The AI Readiness Assessment is the better starting point: 2-3 weeks, fixed price €7,000. All packages and prices are on the pricing page.

In the analysis, I speak to the people who do the work, the process owner, IT, and privacy or compliance where relevant. I look at the data, at any vendor proposals, and at what’s already been tried.

You get a decision document, not a thick report: which use case, which build choice, which conditions for data and privacy, what change is needed, and what the delivery partner needs to deliver to succeed. All the frameworks and analysis are then yours to keep. Want to carry on yourself? The AI implementation step-by-step plan helps with the phases that follow.

More on my background, and why I deliberately don’t sell implementation work myself, is on the about page. Want to start your AI project well prepared? Book an advisory conversation.