Anchoring
Learned, noted, back to business as usual
After the course, everyone knows the basics. Two weeks later the team works exactly as before, because nobody translated the new techniques into their own workflows.
Performance workshop · Enablement architecture
Prompting can be learned in a day. A way of working cannot.
AI at Work is the prosma performance workshop that enables your teams to use generative AI safely and effectively in their daily work: with use cases from your own processes, guardrails that hold, and a transfer plan that changes behavior. It works when your AI licenses have long been in place. And just as well when the rollout is still ahead of you.
2 days on site or 4 remote modules
Up to 16 participants
Transfer included
Sound familiar?
Why AI knowledge so often evaporates
The material itself is quickly taught. What's missing afterwards is the translation into real daily work: Tuesday's email, Thursday's meeting minutes, the research for the client meeting.
Uncertainty
"Am I even allowed to do this?"
Without guardrails, every use is an individual risk: what may go into which tool, what applies to customer data, who is accountable for the result? In doubt, the cautious refrain and the bold do it quietly.
Spread
Two race ahead, the rest wait
Every team has enthusiasts who already do everything with AI, and many who have never opened a prompt window. Leadership stands in between and is expected to explain both.
The starting position in numbers
The competence gap is well documented
0 %
of companies that considered AI and did not adopt it name lack of knowledge as the biggest hurdle
Destatis 2025
0 %
of people in Germany have ever received any AI training at all
KPMG/University of Melbourne 2025
0 %
trust AI outputs. 66 % use them anyway.
KPMG/University of Melbourne 2025
There is a legal side too: since February 2025, Article 4 of the EU AI Act requires providers and deployers of AI systems to ensure a sufficient level of AI literacy among the people operating or using those systems on their behalf. That is one reason. The better one is in the three numbers above.
The workshop
Four modules, one lead-in, your own examples
Compact inputs with continuous practice in between. Not on sample tasks, but on the emails, meeting minutes and research your participants actually have in front of them.
Expectation interviews upfront
15 minutes per personOne to two weeks before the workshop, we speak briefly with every participant. The program follows what is said there, not a standard curriculum.
Understanding and framing AI
Module 1What generative AI can and cannot do. Strengths and limits in a work context, how to classify your own tools, and the role of the human: responsibility, control, quality assurance.
Using AI where it counts
Module 2Typical use cases of knowledge work: texts and emails, summaries and minutes, research and structuring. And the question behind them: where does AI really help, and where only apparently?
Prompting and working with AI
Module 3Role, context, goal, format: the basic logic of good prompts. Plus patterns for recurring tasks, iterative refinement, and handling unclear or faulty results.
Safe and responsible use
Module 4Data protection and confidentiality in daily work, handling sensitive data, typical risks from hallucination to the copy-paste trap. The result: shared guardrails for AI use.
Results
What your team takes away
No slide photos. Six work products that are in use the Monday after.
- The team's use case collection, prioritized by time saved and feasibility
- Role-specific prompt library with templates for recurring tasks
- Quality and safety guardrails: data protection, handling errors, clear responsibilities
- AI policy template for your company, ready to connect to co-determination
- Transfer plan with quick wins, owners and timeline
- Short report with recommendations for scaling and governance
The difference
We measure success in time gained. Not in course satisfaction.
The difference to an AI course rarely lies in the material — that is quickly explained. It lies in the anchoring: use cases from your real processes, guardrails that works council and data protection support, and a transfer that continues after the workshop. On request, we accompany the first weeks with coaching and open office hours until the new techniques stick.
Formats
On site, remote, or with follow-up support
On site
2 consecutive days
At your location, with up to 16 participants. The densest format: one team, two days, shared guardrails at the end.
Remote
4 sessions of 3.5 hours
Four modules over several weeks, with up to 10 participants. Between sessions, participants try out what they learned in their real daily work.
Optional
Transfer coaching
4 to 6 weeks of support after the workshop: office hours, casework on real tasks, refining the prompt library.
No prior knowledge required. Existing experience, for example from private use, is picked up and structured. The goal is a shared, workable level across the team.
Where it fits
The workshop is stage 2 of three: measure, enable, integrate
We support organizations with AI across three stages. The workshop is the second. You can start right here, and it gets more reliable with a measurement first.
Stage 1 · Measure
AI Impact Check
You first want to know what your AI usage achieves today and where enablement is even worth it? The check measures it, in 2 to 4 weeks.
Go to the AI Impact CheckStage 2 · Enable
AI at Work
Your teams should use AI safely and effectively, with shared rules and real use cases. You are here.
Stage 3 · Integrate
AI Integration
Entire work routines and processes need rebuilding? Then a workshop is not enough. The integration consulting continues from there.
Go to AI IntegrationBeyond AI, you can find our other learning formats under Training & Courses.
Common objections
Fair questions, honest answers
Is this a prompt course?
Prompting is one of four modules. The difference lies in the rest: use cases from your processes, shared guardrails, a policy template and a transfer plan. A course teaches knowledge. This workshop changes how your team works.
Our IT can train this internally.
IT rolls out the tools, and that is its job. Changing behavior is a different one: it takes outside facilitation, role-specific use cases and a transfer that reaches beyond the workshop day. The two work best together.
Our people already use AI privately.
A good start, and still different from working with company data: data protection, documentation, responsibilities. The workshop picks up the existing experience and turns it into a shared, regulated practice.
What about the works council and data protection?
Both are considered before anything starts: the guardrails are developed together in the workshop, and the AI policy template is built so co-determination can connect directly.
Do participants need prior knowledge?
No. Little or no experience is fine. Those who already work with AI bring their own examples — that is exactly what we practice on.
My team has no time for this.
Two days are a real investment, no question. They stand against the time currently going into unstructured experiments, duplicated work and unused licenses. In a first call, we run the numbers for your case.