Is this a workshop or a piece of advice?
It is a half day working session. You take home written strategic advice, but the session itself is interactive: we test your use cases live, not behind closed doors.
Stop guessing on an AI stack. In half a day we test Copilot against your use cases, model the business case and write a 90-day roadmap. Written advice included.
Companies that build their AI stack with And AI, from Copilot pilot to custom knowledge assistant.








Four questions every SME asks before locking in an AI stack.
So isn't Copilot a no-brainer? Or are you paying €30 per user per month for something that adds little to your actual processes?
When is that investment defensible? And how much do Dutch WBSO and MIT R&D AI subsidies actually take off the bill?
One vendor for everything feels risky. How do you future-proof this decision without switching stacks every year?
Is that the tool, our usage, or the processes we put it in? Without analysis you'll keep tweaking forever.
Which path fits, we determine during the session based on your processes, data and use cases.
Active in your existing Microsoft stack from day one. Works on the documents, mails and meetings your team already lives in. Our role: calibrate what works and accelerate adoption.
A bespoke knowledge assistant, agent workflow or RAG system on processes Copilot cannot reach. Our role: build, integrate and apply WBSO/MIT R&D AI so the business case closes.
Copilot for broad use, custom AI for the highest-leverage processes. Our role: draw the line so you don't pay twice and don't manage twice.
Five concrete deliverables. No vague guidance, no 6-week discovery.
10-15 pages, ready to share internally with the board or management team. Not a slide deck, a readable document with reasoning.
3-year TCO of Copilot subscription against a custom build (incl. WBSO/MIT R&D AI subsidy impact). One table, one conclusion.
Which processes, which tool, which priority. Not 30 ideas, just the 5-8 that move the needle.
Which steps, in which order, with which expected effects. Executable by your own team or with our guidance.
Within 2 weeks of the session. Questions about implementation, doubts that came up, course corrections. No extra charge.
Four blocks. No filler. From intake to advice, two weeks tops.
We map your Microsoft stack, other tools, processes, data and regulatory constraints. Short questionnaire and a 30-minute call.
We pick 6-8 concrete use cases from your company and stress-test Copilot live. What does it actually solve, where does it fail?
We run the business case: Copilot subscription on 3 years against a custom build with WBSO/MIT R&D AI. Decision matrix on the table.
You receive the written advice, ROI table and roadmap. 30-minute follow-up call within 2 weeks of the session.
Productised offering, transparently priced. Not satisfied, full refund.
Not satisfied? Full refund, no questions. In-company travel costs excluded.
In 2024-2026 we helped SMEs and non-profits with the Copilot-vs-Custom trade-off. Some chose Copilot, some custom, most a hybrid. All with a written, reasoned decision.
And AI is a registered Dutch AI supplier. Clients can apply WBSO and MIT R&D AI to custom-AI projects through us. That cuts the net investment by up to 36%.
We do workshops and bespoke builds. No built-in bias toward one route. See cases Boxing Society, Bolsenbroek and Stichting van de Straat: three organisations, three different paths.
Anthonie Smith and Stefan Frans (co-founders) or an AI Engineer from the team leads the session. No outsourced consultants, no juniors.
It is a half day working session. You take home written strategic advice, but the session itself is interactive: we test your use cases live, not behind closed doors.
You can, and in 30-40% of cases that is also our advice. The problem is that a Copilot pilot without strategic exploration often stalls after 6-8 weeks: you get "it kind of works" but no measurable return. A half-day session prevents 8 weeks of trial and error.
From €15,000 for a first working solution (one bounded process, RAG or agent). With Dutch WBSO and MIT R&D AI subsidies, that can drop net to around €9,500. We run your specific case through the numbers during the session.
Even better. We analyse where it pays off and where it does not, and decide where custom AI should fill the gap without doubling your spend.
Intake within 1 week, session within 2 weeks of intake. Written advice within 5 working days after the session.
One session. One decision. One plan. No surface-level advice, no 6-week discovery.