QA automation engineer in Johannesburg. I test short-term insurance software on SSP Pure, and write most of my automation in Playwright and TypeScript.

About a year and a half into QA. I'm working toward an SDET role, which mostly means getting better at the engineering half of the job: the tooling around the tests, not only the tests.

qa-full-lifecycle — one repository covering the whole testing lifecycle: manual test cases and traceability, SQL data validation, API contract tests, Playwright UI automation, k6 load scripts, and axe-core accessibility scans. Five CI jobs gate it, and every one of them leaves a report behind.

playwright-ai-testgen — generates Playwright specs from acceptance criteria with an LLM, then scores what came back: did it cover every criterion, did it invent a locator, does it compile. Generating the tests is the easy half. The eval harness is the point.

qa-playwright-skills — two Agent Skills. One writes Playwright tests the way I want them written; the other reviews them and won't rubber-stamp anything.

Playwright, TypeScript, Python and pytest, SQL, k6, axe-core, GitHub Actions.

I use AI to draft tests most days. It is quick at a first pass and good at suggesting cases I hadn't thought of. It is bad at knowing which of them matter, and worse at noticing when a test passes for the wrong reason. Most of what I build here is an attempt to close that gap.

maphelatinashe@gmail.com · LinkedIn

Not looking for a role at the moment.