Powering Up QA Teams in the Age of AI – Testing With An Agentic Validation Layer
QA teams shouldn’t pay the price for devs shipping faster with coding agents.
Regression cycles eat days. Every UI change breaks another script. And the tests coding agents write for their own code can’t be trusted, because when those tests fail, the agent often fixes the test instead of the code. So verification lands back on QA.
In this session, QA.tech CEO Daniel Mauno Pettersson shows how QA teams build an agentic validation layer: QA agents that test by user goal instead of by selector, act like a real user, and adapt when flows change. No scripts to rewrite. No code access required.
Includes a live demo of intent based testing, plus what happened at teams that moved their regression runs to agents.
What you’ll be able to do after this session
- Why the bottleneck moved from writing code to verifying it, and why QA absorbed the difference
- How QA agents test user goals instead of selectors, so a UI change doesn’t mean a rewrite
- What to delegate to agents and what to keep for human judgment
- Why independent verification matters when coding agents write and quietly fix their own tests
- How to run cross device tests from a single prompt, with video replay, logs, and qualitative UX feedback instead of just pass or fail
- How agentic tests handle flakiness and reproducibility, and where scripted automation still wins
- What actually changed for teams that moved regression to agents, including where it didn’t work
Learn more about QA.tech here.
About Daniel Mauno Pettersson, CEO of QA.tech
Tech leader and hands-on engineer for the last 20 years. Previously CTO at memmo, Billogram, and Dooer, and CEO at Agigen. Started QA.tech after living the testing problem firsthand while scaling products across markets.
Questions about this workshop
Do the QA agents need access to our source code? No. The agents test by user goal through the interface, not against the codebase.
How is this different from letting a coding agent write its own tests? A coding agent that writes and runs tests for its own code has an incentive problem. When a test fails, it can resolve the failure by changing the test. The session covers why independent verification sits outside that loop.
Does this replace our scripted automation? No. The session covers where agentic testing holds up and where scripted automation is still the better choice.
Will the recording be available? Yes. Register and you’ll get the replay whether or not you attend live.
What does it cost? Nothing. The session is free, brought to you by TestGuild in partnership with QA.tech.
Discover How QA Teams Take Back Control When Devs Ship Daily
Checking your registration status...
© 2026 Test Guild. All rights reserved.