AI training that puts people to work

Since 2 February 2025, article 4 of the AI Act requires organisations using AI systems to ensure "sufficient AI literacy" among employees who work with them. The law does not specify exactly what that means, but it is not a tickbox. The regulator looks at whether your people understand what the tool does, what the risks are, how they verify output, and where they draw the line. For most SMB organisations this means targeted training plus internal guidelines on paper. We help arrange that practically, so you meet the standard without setting up a compliance circus. No standalone e-learning modules nobody acts on, but training that actually puts people to work.

It depends on where you start. An AI basics workshop for a team that has barely used AI takes 1 half-day (3 to 4 hours) including a coffee break. A prompting masterclass for people with some experience is usually 4 hours. For team-specific training we typically schedule 1 day or 2 half-days, with a week in between to practice in the real work. For ongoing coaching we plan for a monthly 1-hour check-in per team. We deliberately keep it compact, because an 8-hour day of classroom training does not work. People learn by doing, not by listening.

Both, and the choice depends on the goal. For groups up to about 12 people onsite (your office or our location in Middelburg) often works better, because hands-on work goes faster in person than in a Zoom call. For distributed teams or follow-up coaching online is fine, and we work with Teams or Zoom plus a shared workbook in Notion or Coda. For some organisations we combine it: kick-off onsite, follow-up online. A practical tip: hands-on labs (where people work on their own use case) almost always run more smoothly onsite, because we can look over the shoulder directly.

Especially for that group, yes. We often see the most valuable participants are people with years of experience in their field who simply have not touched AI tools yet. Their domain knowledge makes them spot what AI can and cannot do in their work much faster than beginners. We do not start with theory or jargon, but with a live demo on a recognisable case from their work. After that they work themselves, with guidance. No question is dismissed. What we avoid: an instructor showing how clever AI is. What we do: let participants experience that they can do it themselves, even without a technical background.

Yes, and that is a different format from team training. An executive briefing takes 1.5 to 2 hours and focuses on decision-making, not skills. We cover: what does AI actually do for your type of organisation, what decisions are coming up in the next 12 months, where do the risks sit on AI Act compliance and privacy, and which investments pay off and which do not. We bring examples from comparable organisations without breaching confidentiality. No sales pitch, no hype, no 80 slides. Often we do this prior to a broader engagement so management sets direction before teams get to work.

A workbook with the cases we covered during training, plus the prompts your participants built themselves. A short set of guidelines for your organisation: what is allowed, what is not, which tools are supported, how to handle client data. Access to a shared Notion or Coda workspace with follow-up material and links to relevant sources. Optionally a recording of the session so colleagues who could not attend can review it. And a tailored list of next steps: not "go use AI", but concretely "these three use cases are ready to roll out next month, these two need more work".

We have participants self-assess before and after on a few concrete points: can I write a good prompt myself, do I know which tools we may use for which type of work, do I recognise when AI output is wrong. We also collect feedback right after training and again after 4 weeks (because that is when you see what stuck). The real signal is not in evaluation forms but in behaviour: do people actually use AI in their work. That is why ongoing coaching helps, because without follow-up about 30 percent of what was learned fades within a month. For organisations that want to report formally on AI literacy (AI Act art. 4) we also deliver documentation of training content and participant list.