AI prototype in 3 weeks
Three weeks, one use case, and an honest answer at the end. I build the prototype for real — working software, not a clickable mockup — and measure whether it's good enough. If it isn't, I'll say so. For a fixed price you'll have avoided sinking a year into it.
There's an AI idea on the table. Nobody can say whether it works, what it costs, or whether it's worth doing.
From AI idea to working prototype — including an honest verdict on whether it should go ahead.
What you get
The yardstick — agreed before building starts
One use case, tightly scoped — straight from the Process scan's shortlist when it exists
A real test set that defines 'good enough', built with the people who do the work today
Working software
The prototype on your data, in your own environment — nothing leaves the house
With citations, so every answer traces back to its source
The verdict
Measured quality against the test set — a number, not a gut feeling
Cost per user per month, worked out for the real thing
A go/no-go with the reasoning attached — and when it's 'don't', what would turn it into a 'go'
Production hardening, SSO, scaling — the prototype proves the case; the production build is its own engagement
Fixing the data foundation the prototype reveals to be missing — that's Your data warehouse
Teaching the wider team to work with AI — that's its own engagement
How it runs
Week 1 — Scope and yardstick
One use case, tightly scoped. And up front: when is it good enough? We build that test set before the first line of code.
Week 2 — Building
Working software on your data, in your own environment. With citations, so every answer traces back to its source.
Week 3 — Measure and judge
Quality against the test set, cost per user per month worked out, and a go/no-go with the reasoning attached.
Where this fits in the path
Frequently asked questions
AI prototype in 3 weeks has a fixed scope and a fixed price: €12,500 excl. VAT. You know exactly where you stand before we start.
The lead time is 3 weeks, with a scope that is fixed up front.
You get: One use case built out into working software; Measured quality against a real test set — not a gut feeling; Cost per user per month, worked out; A go/no-go recommendation — including when that's 'don't'.
Out of scope: Production hardening, SSO, scaling — the prototype proves the case; the production build is its own engagement; Fixing the data foundation the prototype reveals to be missing — that's Your data warehouse; Teaching the wider team to work with AI — that's its own engagement. That is what keeps the price fixed and expectations clear.
Not sure which package fits?
Book a free half-hour call. Then you'll know — and you're committed to nothing.
Book a call