Scope and pricing
We choose the format that fits your question. Scope, price and end date are agreed in writing before I start.
Scan
One week. I look at your pilots and tell you which to kill, scale, or rebuild. Fixed fee. If I say do not build, I will not build it either.
What you get
- Inventory of your pilots
- ROI-potential score per pilot
- Production-risk score per pilot
- Verdict: kill, scale, or rework
- Top 1 to 3 candidates for production
- 90-day plan
Pilot-to-Production
Max four weeks. I ship one workflow to production, at the table with your team and paired with one of your engineers, so the knowledge stays in-house after I leave.
What you get
- One workflow live in production
- Governance and compliance built in
- One of your engineers trained alongside
- Handover in five points: owner, evaluation set, runbook, who steps in, model path
Operating Layer
A retainer with a goal and an end date. I keep the system in production: monitoring, maintenance and the agreements on who owns what.
What you get
- Governance on production
- Maintenance and monitoring
- The operating model your team runs
- A pre-agreed end date
The agreements behind the work
We agree who maintains the software, how we test its quality and what it needs after delivery. If AI adds nothing, I use conventional software or advise against building.
Still comparing? Seven questions to ask any AI agency.AI consultancy or build it yourself?
Build it yourself when you already know which use case you want and someone in-house will own and maintain the system after go-live. Bring in consultancy when that choice is still open, or when your pilots have been stuck in a demo for months. The right route depends on what you already know and who runs it afterwards.
- Build it yourself
- Fits when the use case is settled, an engineer gets time for it and there is an internal owner. A consultant would only be charging you for what you already know.
- A Scan
- Fits when several AI plans are on the table and nobody can say which one makes money. After a week you know, per plan, whether to stop, scale or rebuild, and which one to three candidates are worth production.
- Pilot-to-Production
- Fits when the choice is made but the experience to get it into production is missing. I build one workflow with one of your engineers alongside, so your team can carry on without me.
- Hire someone
- Fits when you have your own data team and enough AI work for years. Then this role belongs on your payroll. My page on forward deployed engineers explains where that line sits.
How an engagement runs
Before I name a price, I look at the material: how many documents there really are, how many exceptions the process has and who touches it. A day of looking prevents a quarter of overrun. Then scope, price and end date go in writing.
In week one we define when something is done: which questions it answers, at what error rate, in which system and judged by whom. In week two I measure how the process runs now, in week three the difference. On day five we decide whether the scope holds. If it turns out more complex, we change the format before the bill grows.
I work at a fixed price because an hourly rate rewards overrun. With AI, an analysis that took two days now takes two hours. With a fixed price you know up front what you pay, and I have every reason to work fast and within scope.
Frequently asked
What does an AI consultant do?
An AI consultant works out where AI makes your business money and what it costs to run that reliably. In practice: choosing use cases, reviewing existing pilots and mapping the route to production, with an owner, a test set and maintenance agreements. With me, the same person then builds the system.
What does AI consultancy cost?
The Scan starts at €1,995 ex VAT for about a week of work. A Pilot-to-Production sprint takes four weeks at most, at a fixed price agreed in writing up front. The Operating Layer is a retainer with a goal and an end date. I do not bill by the hour.
What does an AI implementation cost?
Count on four items: building, hosting, oversight and maintenance. In the Scan I cost those out per pilot and say which pilot is not worth the money. The largest item is rarely the building; it is keeping alive something that should never have reached production.
When do you need AI consultancy?
At two moments: before you build, when you want to know which use case pays off, and when your pilots stay stuck in a demo. In the 2024 BCG survey, 26% of companies were already generating value from AI. That is a figure about companies, not a success rate for individual pilots.
What does responsible AI consultancy cover?
Per use case, the EU AI Act risk tier, because that determines what the law requires; for most mid-market use cases it is limited or minimal risk. In a sprint I deliver an owner, an evaluation set, a runbook and an agreement on who intervenes. The deadlines are on my EU AI Act page.
How is this different from a large consulting firm?
The person you talk to also writes the code. Price and end date are fixed up front, so overrun is my risk. You can check what I built yourself: Signal Match and marktaanpak.nl are live. If you need a team of ten, a large firm fits better; how to weigh that is in my buyer guide.
Where to start
If you know which AI plan deserves production, build it yourself or have it built. If you do not, a one-week Scan is the cheapest way to find out, and only then do you decide whether you need more.
Want to know which AI plan deserves production?
In the Scan I assess your AI work on ROI, risk and production readiness. One week, fixed price. If the answer is no, you get it in writing, with the reason.
30 minutes · free, no obligation