I'm a QA lead, not an AI researcher. Someone handed me the new AI thing and said test it, and I had to work out how.
I have been in testing for twelve years, ISTQB Foundation since 2018. These days I work alongside about twenty other testers in my domain, out of Warsaw. The course is the thing I built once I realised nobody was going to explain AI testing the way I needed it explained.
People with five, eight, ten years behind them, suddenly feeling like juniors again. Every time explainability or model drift came up in a sprint review, we would nod along and then go home and read about it for two hours. Still not really getting it.
Every article online assumed I already knew, or that I would figure it out eventually on my own. Three years of that is long enough.
So I went through a hundred sources myself, and put the answers in one place. Forty-six QA folks went through the free first module and told me where it was still murky.
What is left is the course. Not everything about AI testing, but the specific run of lectures that would have saved me those three years if someone had handed them to me back when this landed on my desk.
Most of the hard ideas in AI testing already have a cinematic parallel, if you go looking for the right one. I lean on them because your brain is already holding the story. I am just attaching the technical part to something you will not forget by Thursday.
A system that slowly stops behaving the way it did on day one, while everyone insists nothing changed.
Catching the real cases against flagging the innocent ones. The trade-off, made uncomfortably concrete.
The original attempt at a test specification for an autonomous system, and every way it quietly fails.
It is not decoration. The analogy gives you a scene to rewind in your head when you are stuck on a question. That is the whole trick. How it is taught matters as much as what is in it.
It took two years and around thirty-five thousand euros to make. Most of that went into getting the explanations right rather than getting them out fast.
The course leads with theory: what these systems actually are, and what words like fairness and drift come to mean once the thing is learning on its own. The ISTQB CT-AI certificate sits on top, for when you want it. You book and pay for the exam separately with ISTQB; my job is making sure you would walk in able to pass it.
The theory is the part that makes people in an AI standup stop and actually listen to you. The paper is nice. Knowing what you are talking about is the point.
I built it short on purpose. One lecture a day, five to fifteen minutes, no filler. If I ever stop being able to teach this honestly, I will stop teaching it.
The first twelve lectures are free. No card, no signup wall. Skip ahead if you already know the basics.