Programme Launch Offer: Save 20% - Book Now

Track Talk, W9

How to Automate Exploratory Testing Using AI and MCPs

Gerard van Engelen

15:30 - 16:15 CET, Wednesday 17th June

Exploratory testing uncovers real bugs that scripted tests miss, but it’s manual, inconsistent, and hard to share across teams. This talk shows you how to capture that intuition in a repeatable framework using AI agents and the Model Context Protocol.

You’ll see how to encode your testing strategy as system instructions: goals, heuristics, and exit conditions. We’ll use basic tooling that is available to everyone, and set up a framework that can be used for automating your exploratory testing efforts.

We’ll run the same test charter in two contexts: interactively during development for fast feedback, and headless in CI alongside your existing test suite, complete with screenshots and standard reports. And because the proof of the pudding is in eating it, I’ll do it all live on stage—so be prepared for some live demo carnage.

You’ll leave with a working reference implementation, including Playwright configuration and a GitHub Action you can integrate immediately.