Executive committees
Understand what generative AI changes, its limits and its risks, so you can decide how it is used, set the rules and discuss it with staff without hand-waving.
AI training
for teams who use it well.
Our AI training teaches your staff to use generative AI on their own work: drafting, summarising, searching and analysing, while knowing how to check what the tool produces. Every workshop is built on your documents and internal rules, on site or remotely, from the executive committee to frontline teams.
Learning to use it, and when to doubt it.
Useful AI training goes well beyond a tool demo. It shows how to phrase a request, supply the right context, review an answer and catch a plausible-sounding mistake. It also covers what must never be handed to an assistant: personal data, client information, strategic documents. Exercises are drawn from your own business, not from generic examples.
Many companies hand out licences and consider the matter closed. A few curious people make progress, the rest give up after disappointing attempts, and everyone invents their own rules. Another pitfall is a lecture-style session with no practice, soon forgotten. The biggest risk remains overconfidence, when someone reuses an AI-generated text or figure as is, without checking it.
Our training is aimed at executive committees who want to understand before deciding, at admin, sales and support teams who produce a lot of written work, and at managers responsible for enforcing usage rules. Content is adjusted to your sector and to each person’s comfort level, from complete beginners to regular users who want to go further.
Who we train.
Corporate AI training does not cover the same ground for an executive committee as for a support team.
Teams who write a lot
Administration, sales, HR, customer support, communications: drafting, summarising, rewording and replying, with checking methods suited to each role.
Managers
Supporting adoption in their department, enforcing the acceptable use policy, spotting good practice worth spreading and drift worth correcting before it takes hold.
AI champions
The people who will become their colleagues’ first port of call: deeper work on complex cases, designing reusable instructions and keeping an eye on the tools.
Four training formats.
From the boardroom to the desk, everyone gets what they need.
Leadership session
What generative AI can do, where it falls short, the legal and reputational risks, and the questions to settle before rolling it out more widely.
Role-based workshops
Sales, HR, finance, support or production: each group practises on its own document templates and everyday requests.
Acceptable use policy
Approved tools, forbidden data, mandatory review: a short policy, written with you and given to every participant once the workshops are over.
Coaching for champions
One-to-one sessions for the people who will go on to support their colleagues, based on real situations met in their department.
What every participant takes away.
Training leaves a written trail, or it is forgotten within weeks.
- Full materials, handed over after every session.
- Exercises built on your anonymised documents, reusable in-house.
- A library of standard requests by role, tested during the workshops.
- A generative AI acceptable use policy signed off by leadership.
- A checklist for reviewing answers: sources, figures, tone, sensitive data.
- An end-of-programme review with the needs identified and suggested next steps.
- A follow-up check-in to clear blockers that surface after the workshops.
Training so the tool sticks.
A new tool is only worth something once the people unused to it adopt it. For the digitalisation of judicial procedures in the Canton of Fribourg, Victor Michel, the studio’s founder, trained users with little digital experience, ran the acceptance sessions and prepared the move from paper to the tool. Our AI training follows the same change management: start from real tasks and support people until they are self-sufficient.
Canton de Fribourg · GovTech · Digital justice · 2023–2024
- GovTech
- digital justice in Switzerland
- BDD
- requirements turned into testable specifications
The training journey.
Good training takes as much preparation as delivery.
- 01
Needs assessment
Conversations with a few participants and their managers to pinpoint the tasks to work on and the tools already in use.
- 02
Exercise design
Materials and case studies built from your anonymised documents, pitched at the level of each group.
- 03
Workshops
Little theory, lots of practice: everyone leaves with ways of prompting tested on their own work.
- 04
Embedding
Full materials handed over, usage policy signed off and a follow-up check-in to clear any blockers that surface after the sessions.
AI training: what really works.
ChatGPT, Copilot, Mistral, Gemini: the tool matters less than how people use it day to day.
Prompt engineering is not a magic formula
Lists of “perfect prompts” are everywhere and age quickly. What lasts are principles: give context, specify the expected format, provide an example, break a long task into steps, ask the tool to spell out its assumptions. Our workshops teach these habits on each participant’s real work, so they hold up with today’s assistant as well as next year’s, whichever provider your company ends up choosing.
Checking is part of the job
Generative AI produces fluent text, including when it is wrong. Serious training teaches people to spot plausible errors: an invented figure, a non-existent reference, a misquoted rule, a tone unsuited to the client. It also sets the level of review according to the stakes: quick for an internal draft, careful for a document that commits the company or goes out to a customer.
ChatGPT, Copilot or an in-house tool?
The right assistant depends on your office software, the confidentiality required and the intended uses. An assistant built into your email and documents offers neither the same possibilities nor the same guarantees as a standalone tool. We train on the tool your company approves; if no choice has been made, the leadership session helps settle it on clear criteria rather than fashion.
The AI Act and staff competence
Since February 2025, Article 4 of the EU AI regulation, softened in 2026 by the Digital Omnibus, has asked organisations that use AI systems to take measures for their staff’s AI literacy. Beyond the obligation, it is common sense: a trained team makes fewer mistakes and protects data better. The materials handed over and the end-of-programme review keep a record of the steps taken; the precise legal reading remains a matter for your advisers.
Measure take-up, not satisfaction
A feedback form on the day mostly measures the mood in the room. What counts is usage a few weeks later: who is using it, on which tasks, with what difficulties. That is what the follow-up check-in after the workshops is for. It lets us adjust the policy, extend the library of standard requests and identify topics that deserve deeper work or a more structured project.
Tailored, off-the-shelf or self-taught?
Each format has its strengths; they do not meet the same need.
| Criterion | Brutux Studio | Catalogue course | Online tutorials |
|---|---|---|---|
| Exercises | On your documents and your tasks | Generic, cross-sector cases | Examples chosen by the author |
| Internal rules | Usage policy written with you | General reminders | None |
| Groups | Formed by role and by level | Participants from several companies | Individual learning |
| Location | At your premises or remotely | Training centre or virtual classroom | Wherever and whenever you like |
| Follow-up | Check-in after workshops, coaching for champions | Rarely included | None |
| Link to strategy | Can sit within a CRAFT engagement | Independent of your projects | None |
Frequently asked questions.
AI training for teams: answers to the questions we are asked.
What level do you need for AI training?
No technical prerequisites. We form groups of similar levels and adapt the exercises: beginners learn the basics of prompting, more advanced users work on breaking down long tasks and verifying sources.
Which tools do you use during the workshops?
The ones your company allows, or the ones it is considering. If nothing has been decided yet, we present the main families of assistants and the criteria for choosing: confidentiality, hosting, integration with your office software.
Are the workshops held at our premises or online?
Both. In Haute-Savoie, Geneva or Lausanne, we come to your offices, which makes it easier to practise on your own workstations. For distributed teams, shorter remote sessions work well.
How do we keep confidential data out of AI tools?
Through simple rules everyone understands: what can be pasted into an assistant, what is off-limits, which accounts to use. The workshops put them into practice on concrete cases, and the policy handed over afterwards sets them down in writing.
Is training our teams enough to adopt AI?
It is an essential first step, but rarely enough on its own. Without priority use cases and governance, practices stay scattered. Training can therefore sit within a broader consulting engagement, or come before it to get people ready.
Can AI training be funded by a French OPCO?
Funding depends on your OPCO, the size of your company and the certification of the organisation delivering the training. We cannot guarantee it in advance: check with your OPCO before committing, and raise the question during our first call so we can look together at what applies in your case.
How long does AI training take?
It depends on the audience and the goals. A short session is enough for leadership to lay the groundwork; teams progress better with several spaced-out workshops that leave time to practise in between. The exact duration is set out in the proposal, after the needs assessment.
Should we train on ChatGPT or Microsoft Copilot?
On the tool your company approves. The principles of prompting and checking are the same; what differs is access to your documents, confidentiality guarantees and integration with your software. If you are still undecided, the leadership session compares the options against your criteria.
Does the AI Act require us to train our staff?
No standard training is imposed, but Article 4 expects organisations that use AI to take suitable measures for their staff’s AI literacy; the 2026 Digital Omnibus softened this point. Training your teams goes in that direction. To find out exactly what the text requires in your situation, talk to your lawyers.
Do you offer prompt engineering training?
Yes, built into the role-based workshops rather than as an isolated theory module. Participants learn to structure a request, supply context, iterate and build reusable instructions, on their own tasks. Champions go further, as far as configuring dedicated assistants for their department.
(Let’s work together)
Let’s talk about your project.
30 minutes to talk about your project, in Thonon or by video call. I’ll tell you honestly what will make the difference.
Book 30 minutes ↗