A four-figure check prevents a five-figure mistake.
Before you put €15,000 to €50,000 into an AI project, I spend 1–2 weeks with your real data and systems to find out whether it holds — and commit to a verdict in writing. If you build afterwards, the fee is credited in full. If the verdict is “don’t build this”, the check just saved you a multiple of its price.
1 week · 1 use case
- Two-hour kickoff workshop, remote or on site
- Analysis of one use case on your real data and systems
- Review of data availability and quality
- Solution architecture sketch with an operating-cost estimate
- Go or no-go verdict — in writing, with reasons
- Report plus a 90-minute results session
2 weeks · up to 3 use cases
- Everything in Compact — for up to three use cases, ranked by value
- Review of your data and infrastructure
- Deployment assessment: cloud API, EU cloud or your own servers — with a GDPR read
- Make-or-buy comparison where ready-made tools exist
- Implementation roadmap
- A binding fixed-price offer for the build
A report you can base a budget decision on
Not a slide deck: a document leadership reads and the IT team can verify.
Because you are buying a verdict, not a sales call
The free intro call sorts your idea — gladly and without obligation. The analysis behind it is engineering: I examine your actual data and systems and commit myself in writing. Because you pay for it, “don’t build this” is an admissible answer — and exactly that is what makes a “go” carry weight.
A good fit if …
- You have a concrete process in mind — order intake, document review, support, reporting.
- Your data lives in real systems: ERP, document management, email, databases.
- You want a production system — and can take a reasoned “no” if that is the answer.
Not a fit if …
- You are looking for a free proof of concept, or slides for the next board meeting.
- You need a large team in-house for many months — projects that don’t fit into one engineer’s hands get a reasoned no.
- You want AI for the press release rather than for a process.
Confidentiality: on request I sign an NDA before the kickoff. Everything else procurement and IT need is under Procurement & IT.
