01 — Format
Hands-on instead of slide marathons
We have seen enough AI trainings that consist of 90 slides of theory and one live demo of ChatGPT. You know the result: polite interest, then everything stays the way it was. Our workshops flip that around. Every participant brings a real work task from the current week, and that is exactly what we work on. A proposal that needs writing. Meeting minutes that need summarizing. An Excel analysis that eats an hour every single time.
That way everyone in the room immediately sees what AI is good for in their own work and where its limits are. We have been building with language models ourselves since the first GPT-3 APIs, after twenty years of digital projects before that. So you learn from people who use this stuff productively every day, not from trainers with memorized slides. The training is one building block of our work as an AI agency: whatever surfaces as a use case in the workshop, we can also build afterwards.
02 — Current
What's changing right now: training is now mandatory
Since February 2025, Article 4 of the EU AI Act requires every company that uses AI to have demonstrable AI literacy among its staff. In 2026 this gets serious: the supervisory structures are in place, and with the next stage of the regulation taking effect in August 2026, the literacy duty moves into the focus of audits. If you have nothing to show by then, you have a problem that goes beyond an embarrassing audit, because when damage results from improper AI use, the company is liable for breached duties of care.
The good news: the AI Act does not prescribe a fixed curriculum. Trainings, internal policies and multiplier programs are all recognized; what matters is documenting the measures. That is exactly what our workshops deliver: attendance records, a content overview and the AI policy that comes out of them.
Looked at soberly: the training duty is that rare case where compliance and self-interest want the same thing. A team that handles AI safely is at the same time your best protection against data incidents and your biggest productivity lever.
03 — Module
Prompt fundamentals for the whole team
The entry point for everyone: half a day, after which everybody in the company understands how language models work, why they hallucinate and how to phrase tasks so that usable results come out. We practice giving context, defining roles, providing examples and critically checking results. And we settle the question that runs beneath the surface in every company: which data is allowed into which tool in the first place?
04 — Module
Deep dives for departments
After the fundamentals it gets specific, because accounting and marketing need completely different things. In marketing it is about content production, campaign research and reporting, often directly connected to our performance marketing. In sales it is proposal drafts, meeting prep and CRM upkeep. In HR it is job ads, application screening and the delicate questions around fairness and data protection. In accounting it is invoice checks and reconciliations.
05 — Module
Use case identification in the workshop
The most valuable side effect of every training: the best AI ideas do not come from us, they come from your team. The person who checks invoices every day knows best what is annoying about it. We moderate this process in a structured way, collect the candidates and rate them together by effort and impact. How that list turns into an implementation plan is what we describe on our page about AI for SMBs.
Want to know upfront whether a workshop is worth it for you? Drop us a line and we will look at your starting point together in a no-strings first call and tell you honestly which format makes sense.
06 — Outcome
Internal AI guidelines as the outcome
Every training of ours ends with a document that lasts: your internal AI policy. It records which tools are approved, which data may go where, who reviews new tools and how AI-generated content is handled. Not 40 pages of legalese, but two to three pages that everyone reads and understands. If you run a platform like Langdock, we anchor the policy directly in the tool configuration so the rules exist in practice, not just on paper.
07 — Staying on it
Follow-up sessions instead of a one-off effect
Four to six weeks after the workshop we come back, remote or on site. That is when you see what stuck: which prompts the team actually uses, where things get stuck, which new questions have come up. These sessions are short, concrete and often the moment when tentative first attempts turn into solid routines. On request we train internal multipliers who onboard new colleagues themselves.
How fluent is your team really?
Tell us briefly how big your team is and which tools you already use today. We will suggest a training format that fits you. First call free of obligation, 30 minutes.