Learn where AI goes wrong and how to test it. Data, model quality and testing methods, brought together in one clear course.
At your own pace · also on Telegram
For testers who want to work confidently with AI features. Start with AI fundamentals and build towards choosing a testing approach.
All 12 modules →A difficult topic does not have to stop your learning. Kai helps you work through the course material.
Sign in to save your progress.
How do I test image recognition if I don’t know every correct answer?
Start with a change that should leave the answer the same. Make a cat photo slightly brighter: it is still a cat. If the model changes its answer, investigate. This lets you check behaviour without knowing the answer for every image.
Voice generated with AI.
A short explanation of one topic, with examples from AI systems.
See your result and understand why the answer is right.
If something is unclear, ask for a simpler explanation or another example.
I’m a QA Tech Lead. This course grew from questions I encountered while working with AI.
Over two years, I brought the theory into a structured course, so you can understand what each test checks and why it matters.
More about the author and course →Get to know the course for free. The full programme is one payment.
If you have a foundation in testing and want to understand AI, yes. No machine-learning experience is needed: we start with the basics. The course teaches theory and checks your understanding through questions; building your own project is not part of the programme.
The first 12 lectures introduce the course and AI fundamentals, with quizzes and 3 questions a day for Kai. Sign in to save your progress.
In your browser or Telegram. Course materials are in English, Russian and Ukrainian. Polish and Korean interfaces are available; some materials in those interfaces are shown in English.
Full access lasts 6 months from purchase, with no subscription. Lectures open as you progress, and you can revisit completed topics. There is a 14-day money-back guarantee. Purchase terms.
The course helps you prepare for CT-AI v1.0. Certification is optional: the external exam is taken and paid for separately and requires a CTFL certificate. Our course certificate does not replace ISTQB certification.
Version 2.0 is already available. The v1.0 exam is available until 21 April 2027 in English and 21 October 2027 in other languages. Check the current ISTQB requirements before booking.
