Why we run corporate courses as modules
A one-day course gets people started, and for many that is exactly what is needed. If a whole department is going to change how it works, a modular programme can be the better model.
Vibelabs5 min read
The reason lies in what happens afterwards. New methods compete with people's own projects and their own deadlines the moment the course ends. What is not tried in the first few weeks is rarely tried at all, and the skill pools with the most eager while the rest of the department carries on as before.
A modular programme is built to keep it going over time.
What a modular programme is
The training is split into several sessions spread over a few months, with real work in between. Each module has its own theme and builds on the last. Between modules the participants apply what they have learned to their own tasks, and we adjust the next module according to what actually worked.
A typical programme looks like this:
Preparatory workshop, two hours. We start with the department's leaders, before any modules are set. We map which tools and systems you actually work in, and pick the cases the modules will use. That hour is what makes the content yours, rather than a generic course with your logo on it.
Three modules, three hours each. A few weeks apart. Each module has a theory part and a practical part, where participants work on tasks from their own week.
One hour of evaluation after each module. With the leaders, not the whole group.
The evaluation is the most important hour
That hour between modules is easy to cut, and it is the one we hold on to hardest.
It is where we find out what people actually did with what they learned. What was adopted straight away, what was tried and dropped, and where it stalled. Often the obstacle turns out to be something other than the tool: an unclear policy, a system nobody has updated in a year, or a workflow where two roles are waiting on each other.
Then we change the next module. A course planned out in detail before it starts cannot do that.
What the modules cover
The themes are assembled around what the department needs, but the order usually follows the same logic.
The first module is about context: the difference between ad-hoc prompting and systematic context building. What the tool knows about the task governs what you get out of it. This is the ground everything else rests on, and it is also where most people are leaving the most behind.
The second is about shared standards. Best practices, skills and setups the team owns together, so the method does not live in one person's head. Here we build something concrete you take with you.
The third is about workflow and handover. How AI changes the transitions between roles, and what the next person needs in order to carry on. This is usually where the gain is largest, and where it most often goes unclaimed.
The content sits on the tools you already use. No new stack has to be introduced for the course to make sense.
When the modular programme is not right
The modular programme is not the answer to everything, and it asks for more time than a one-day course.
If a handful of people need to get going quickly, the one-day course is the shortest way in. If you need something actually built, that is a different service, and we are happy to do both.
And if you do not have the capacity to do any work between the modules, the whole point falls away. Then you get three disconnected sessions instead of a programme. We say so in the preparatory workshop, not afterwards.
The experience behind it
Over 2,000 people from 200+ companies have been on our courses, and 95% report that they are satisfied. We have run over 100 events.
What we have learned from that is that the format matters more than the content. The same three hours give very different returns depending on whether participants get to use them on something real in the meantime.
Get in touch
If you are raising AI skills across a whole department, we are glad to talk about how a programme could be set up with you. We go through what you work on today, and we are honest about which format fits. Send us a message and we will find a time.
- AnalysisFrom vibecoding to loops to graphsSix weeks. That is how long it took to get from “stop prompting, design loops” to “are we still talking about loops?”. Here are all three steps, and where you should be standing.
- Case studyHow AI agents changed the running of AtelieAtelie is an art platform with over 1,000 professional artists and 10,000 works. It was built by senior developers over several years. Today I run it alone. Here is what actually changed.