QA Automation Engineer — TypeScript, Playwright, CodeceptJS.

I own E2E automation and QA tooling at Testomat.io: a CodeceptJS 4 + Playwright + TypeScript framework rebuilt from scratch — 70+ page objects, 300+ scenarios, 4 parallel workers, one codebase for three environments — and the CI around it (nightly two-env matrix, PR gates, run queueing). Earlier: cut CI regression from 5h 27m to 2h 31m (-54%) on a 500+ test suite with Playwright sharding in GitHub Actions.

7 merged bug fixes in CodeceptJS core — parallel run-workers, retry logic, the Playwright helper, reporting plugins:

• #5572 — recorder: stop mutating shared defaultRetryOptions
• #5571 — retryFailedStep: exact-name matching for ignored steps
• #5566 — analyze: AI SDK ImagePart format for vision screenshots
• #5564 — step: rename Step.name to Step.title for consistency
• #5561 — workers: do not exit worker on unhandled rejections
• #5552 — playwright: reset cleanup state to prevent worker crashes
• #5508 — stepByStepReport: cheerio ESM import and missing screenshots

Public functional audits from that work: codeceptjs-retry-test · codeceptjs-plugins-test · codeceptjs-sharding-sandbox

• Test-case pipeline on Claude Code skills: explores the app, writes user-scenario test cases, self-reviews them through an LLM agent plus quality gates, publishes to the TMS
• A second skill that covers those cases with autotests — one verified test at a time
• Explorbot (AI exploratory bot) daily in CI with LLM-triaged reports

Author of vyvchy.com — a free QA learning platform in Ukrainian: 14 sections of theory with real-world examples and interview pitfalls, quizzes with explained answers, flashcards for pre-interview review, diagnostics, and an AI mock-interview tool.

CV and contacts: djinni profile