Or: How I wasn’t allowed to look out of the window for three hours… “Failure to comply with these rules of conduct will be regarded as an attempt to cheat and will result in a fail.” This is how my journey to becoming an AI manager began.
Before a single exam question even appears, I read through a list of requirements set by the Fraunhofer Institute. Audio, video and screen data will be analysed. Exam invigilators from the Fraunhofer Personnel Certification Body can virtually look over my shoulder. No headset. No notes. No other devices. No second person in the room. No documents within reach. No talking to myself.
Talking to myself… I stare at this spot for a little longer than the others.
My AI Manager exam
I’d brought along some nuts and apple slices. After all, ‘unpackaged food’ is allowed. Drinks too. You’re allowed to leave the room once to go to the toilet. Nothing more.
My shelf in the background? Covered with cloths.
The second screen? Turned round and covered as well.
All documents? Removed from the room.
Smartphone? Not nearby, of course.
Paper? Forbidden. Taking notes? Not allowed.
But what’s really on my mind is this: I tend to speak quietly to myself when I’m thinking. Especially when I’m organising my thoughts or working through complex concepts. That’s exactly what’s forbidden. No talking to yourself. No reading questions aloud. No using your voice to organise your thoughts.
And then there’s this: don’t disappear from the camera’s view.
I usually look out of the window when I’m thinking. Always. That’s my focal point. My way of sorting through information.
This time, the task is: three hours staring at the screen.
Three hours. No multiple choice. Just application.
As the exam begins, it quickly becomes clear: this is no tick-box test. No multiple-choice questions that can be answered on gut instinct, but challenging, open-ended questions. Transfer. Application. Strategy. Architecture. Governance.
And then they’re all there: the last three weeks.
- AI Basics.
- AI Ideation.
- Strategising.
- AI Design.
- AI Implementation.
Three weeks between the familiar and the ‘tough stuff’
The course itself was delivered entirely online. Three intensive weeks with a clear structure: kick-off webinars in the morning, self-study modules in the late morning, and interactive workshops in the afternoon. There were also optional in-depth modules and, at the end, this exam.
AI Basics
A solid foundation. What is AI really – and what isn’t it?
How do automation, data analysis and machine learning differ?
Less buzzword bingo, more clarity.
AI Ideation
Much of it felt familiar: developing visions, identifying use cases, prioritising, and fleshing them out.
And yet it was helpful to work through everything systematically, including the AI Service Canvas and clear assessment against strategic objectives.
Strategizing
This is where things became significantly more challenging.
How does AI specifically contribute to business objectives?
How do you establish governance structures?
What roles and skills are required?
How do you meaningfully embed ethics, risk management and regulatory requirements such as the EU AI Act within the organisation?
By now, at the very latest, it becomes clear: AI is not an IT project. It is organisational development.
AI Design: my personal ‘moment of respect’
Data strategies. Reference architectures. Cloud vs. Edge. Technology. Human-AI interaction. UX issues. Data quality as a strategic asset.
That was really ‘heavy stuff’. And at the same time incredibly valuable, because it becomes clear: good AI doesn’t just emerge from the model, but from the seamless integration of data, architecture, user-centricity and governance.
AI Implementation
From proof of concept to production.
ML lifecycle. MLOps. Monitoring. Dealing with data drift. Scaling roadmaps.
The module that determines whether AI projects remain at the pilot stage or achieve real effectiveness.
But now back to the AI Manager exam
At some point, I forget about the camera. Forget that I’m not allowed to look out of the window. Forget that I’d normally be talking quietly to myself. I think. I write. I organise my thoughts – this time, only in my head. Three hours of concentration. A short break. Nothing more.
And then: submission.
A few days later, the news.
I’ve passed!
Now I’m officially an AI manager
With the certificate from the Fraunhofer Personnel Certification Body, I’m now officially an AI manager. What has changed for me?
I have greater clarity, more methodological confidence and an even deeper understanding of how strategy, technology, organisation and governance must work together for AI to be truly supportive.
And above all, I’m even more eager to put this knowledge into practice.
If you or your company are considering introducing AI, prioritising use cases, establishing governance or strategically developing existing initiatives – let’s talk.
I look forward to our first discussion.
Warm regards
