AI training
that starts from your processes
Not a course catalogue, but the step that makes the next project possible. Taught by the people who build AI systems in production, not by people who have read about them.
AI projects stall on people, not on technology
The script repeats itself: management decides to invest in AI, a pilot starts, the pilot works, and then nothing happens. Not because the technology disappointed, but because nobody inside the company can say precisely which processes are worth touching, what can be delegated to a model and what cannot, where automation ends and a human signature has to remain.
That is a skills gap, and a generic course on what an LLM is will not close it. It closes by working on the processes you actually run, with the people who own them.
Assembled around where the team actually starts from, not around a fixed syllabus
Getting started with AI
For those starting from zero who have to decide. What models can do today, what they cannot, what it costs and what risks it carries: enough to stop choosing by hearsay.
Closing the technical gap
For development teams about to work on agentic systems: RAG, orchestration, guardrails, evaluating results. With the code, not with slides.
Vertical programmes on one process
We take one of your processes (quoting, support, audit, reporting) and work on that: what to automate, with which data, under which control.
Governance and the AI Act for decision makers
For management, legal and compliance: which obligations fall on those who use AI, what has to become engineered controls inside the software, and what a document cannot solve.
Working alongside on the project
The format that works best: we build the first system together and your team learns by doing it, instead of receiving a delivery to maintain blind.
Handover
At the end of a project the team has to be able to maintain the system on its own. It is what makes code ownership real, and it is part of every delivery.
Why you will not find a course catalogue
A catalogue with titles, durations and prices would be easier to read, and would serve little purpose. Two companies asking for the same thing almost always start from different places: one has clean data and a team standing still, the other has developers ready and nobody able to decide which processes to touch. The same course would be wasted on one of the two.
That is why training here is not a separate product: it is the step that precedes or accompanies a project, and it is sized against it. If the project turns out not to make sense after the programme, we tell you.
Taught by the people who build
The people running these programmes are the same ones who design and ship the agentic systems they talk about: multi-tenant platforms, agents with guardrails and human-in-the-loop, integrations inside existing ERP and CRM systems. The examples come from there, including the ones that went wrong.
What companies ask before starting a programme
You are not a training company: why train with you?
How long does a programme last?
Does the team need to know how to code?
Do you run training without a project together?
How do you tell whether it worked?
How do you pick the point to start from?
Tell us where your team is starting from, what the project is and where you want to get to. We will give you the fixed points to anchor your climb to.
