QA engineer · 15 years · building QALabs

I find the bugs your test plan never thought to look for.

15 years building and breaking software. I design test strategies, automate them, and build tools that generate the tests for you.

Playwright · TypeScript · Cypress · k6This site's own test report

How QALabs works

Not a prompt wrapper. The model only writes at stage three — everything before and after it is deterministic.

01
Crawl
Playwright walks the app and captures the accessibility tree at every reachable state.
02
Reduce
The route graph collapses into flows — entry point to state change.
03
Generate & verify
Risk-rated cases per flow, then a second pass drops anything the crawler never saw.

Turns a URL into a reviewed, risk-rated test suite in under 10 minutes.

An AI agent that crawls a web app and writes the test cases — then turns them into Playwright or API scripts.

Next.jsPlaywrightClaude Agent SDKPostgresDrizzle

Cut the nightly regression run from 6 hours to 11 minutes and flake to under 1%.

Inherited a 2,400-test suite with a 40% flake rate that nobody trusted. Rebuilt it into a gate the team actually respected.

PlaywrightTypeScriptGitHub ActionsDocker

Sep 1, 20261 min readFeatured

Most "AI writes your tests" demos fall apart on real apps. Here is the architecture I landed on after the naive version failed.

#ai#qa#architecture

Aug 14, 20261 min read

The bugs that hurt are never the ones in the test plan. Some patterns I keep seeing across teams, products, and decades.

#qa#career#process

I find the bugs your test plan never thought to look for.

15 years building and breaking software. I design test strategies, automate them, and build tools that generate the tests for you.

Playwright · TypeScript · Cypress · k6This site's own test report