QA Ark is picking a fight

Not with a person — with a vague idea. The idea that quality engineering can carry on as normal while AI rewrites how software gets built around it.

It can't, and most of us know it can't. But the advice on what to do about it comes overwhelmingly from two camps: vendors selling you the answer, and pundits who haven't shipped a test suite in years. There's a gap for the boring-but-true version — what actually works, measured, from inside a real engineering org. That's the gap this newsletter is going to occupy.

What's changing

From this issue on, QA Ark is about quality engineering for the AI era. Concretely, that means three threads:

• AI-era test strategy — where AI genuinely earns a place in your quality workflow, and where it's expensive theatre
• Automation at scale — suite economics, rot, flake, and the metrics that predict trouble before it hurts
• Team enablement — who owns quality, how skills need to shift, and how to bring a team along rather than leave it behind

What's not changing: no fluff, no filler, no engagement-bait hot takes about AI replacing testers. If a claim shows up here, it comes with reasoning you can check against your own suite.

The Playbook

The strategy essays now live in one place, in reading order: The AI-Era Quality Playbook.

Nine chapters, from ownership through suite economics to AI in practice. Each stands alone, but the argument builds: quality is everyone's job, automation should earn its keep like any other engineering investment, and AI belongs in the workflow where it removes toil — not where it adds noise.

If you read one chapter this week, make it The Cost of a Slow Test Suite. It's the piece I find myself sending to engineering leaders most often.

One question to take into your week

"What did our test suite cost us this sprint — and what did it catch?"

Not rhetorically. Actually tally it: execution minutes, flake-triage hours, the PRs that sat waiting on a red build that turned out to be noise. Then put it next to the defects the suite genuinely stopped. Most teams have never seen those two numbers side by side, and the reaction when they do is where every good test strategy conversation starts.

If you run the numbers and they're uncomfortable, chapter three of the playbook is the follow-up.

From the workbench

Two tools where these ideas get pressure-tested:

• Playwright Smart Reporter — AI-powered failure analysis and flaky detection for Playwright. The suite-rot metric from the playbook started life here.
• Stagewright — turning raw test results into the dashboards those uncomfortable conversations need.

Stay connected

• Join the Discord community — the suite-cost question above makes a good first post
• Want this thinking applied to your team? Workshops are open for registration

More chapters land here first. See you in the next issue.

Gary Parker Quality Architect & Strategic Advisor