AI literacy for people who already use AI
Your team is already pasting things into chatbots. This is the day where they find out how those tools actually work, where they fail, and which of your data must never go near them.
Two formats
Half day
€490
About 4 hours
The essentials, for a whole team at once. How these models work, what they are bad at, what must never be pasted into them, and how to get better answers out of them. Ends with the usage policy discussion.
Two days, hands-on
€1,490
2 × full day
The same ground, then people work on their own actual tasks with us in the room. They leave with prompts and small workflows built for their own job, which is the part that survives contact with a Monday morning.
What people learn
Four things, in this order, because the fourth one does not stick without the first three.
- How the tools they use actually work — enough of the mechanism to predict behaviour: why it invents a plausible citation, why it is confident when wrong, why the same question gives two answers.
- Where they go wrong — hallucination, stale knowledge, arithmetic, being talked into things, and the specific failure modes of the tools your team actually has licences for.
- Which data must never go in — worked through against your own categories — personal data, client material under NDA, unpublished figures, credentials — rather than in the abstract.
- How to use them well — on real tasks from the room. Not prompt tricks that stop working next quarter, but the habits that hold: give context, ask for the reasoning, check the claim, keep the human decision.
What you get afterwards
- An AI usage policy adapted to your company — not a template with your name on it: written around the tools your people actually use and the data you actually hold. Yours to adopt, edit or ignore.
- Attendance records — who attended, what was covered, on what date, signed. This is the part that matters if anyone ever asks what you did.
- The materials — slides, the worked examples, and the prompts built on the day.
- A follow-up question window — a month of email questions from attendees, because the real questions arrive a fortnight later.
The EU AI Act angle
Article 4 of the EU AI Act, as amended by Regulation (EU) 2026/1744 and in force since 27 July 2026, requires providers and deployers of AI systems to take measures to support the AI literacy of their staff. It does not prescribe a particular level, a syllabus or a certificate.
A workshop with a documented syllabus, an attendance record and a written usage policy is a practical and evidenced way of taking such a measure. That is the whole of the claim we make about it.
This is not legal advice, and we are not the people to tell you what your obligations are. If you want a legal opinion on your situation, ask a lawyer.
Questions we get asked
How many people can attend?
The half day works up to about 20; beyond that it stops being a conversation. The hands-on version works best at 8–12, because everyone gets time on their own task.
Do they need to be technical?
No, and mixed groups work better than uniform ones. The people who know the work best are usually not the technical ones.
What if our team is sceptical about AI?
Good. The sceptics ask the useful questions, and a large part of the day is about what these tools cannot do. Nobody leaves having been sold anything.
Online or in person?
Both work. The half day is fine online. For the two-day version, in person is noticeably better — the value is in looking over someone's shoulder.
Which language?
Hungarian, Serbian or English, including the materials and the policy. Mixed-language groups are fine; we have never yet run a room where that was the problem.
Is travel extra?
Not in Vojvodina or Hungary. Further afield, we will tell you the travel cost before you book.
Book 20 minutes first
Tell us which tools your team already has and what they are using them for, and we will tell you which format fits — and whether you need us at all.