I automate the things that break at 3am — APIs, data pipelines, CI gates.
4+ years shipping stable test suites for production backends.
| Area | What I actually do |
|---|---|
| API automation | REST, SOAP, gRPC — Pytest + Requests, end-to-end contract validation |
| Data validation | PostgreSQL & Redis assertions, Pydantic response typing with strict mode |
| CI/CD quality gates | Blocking jobs on failure, pytest-xdist parallel runs (2–3× faster) |
| Reporting | Allure dashboards readable by devs, QA leads, and PMs |
| Stability | Retry strategies, flaky test analysis — +30% pass rate on volatile environments |
- 🚀 2–3× faster test suite execution via parallel runs with pytest-xdist
- 🛡️ Zero broken builds reaching staging — CI gates catch regressions at the PR level
- 📈 +30% test stability through intelligent retry logic and flaky test isolation
- 📋 Standardised test case templates adopted across the QA team
PDF requirements → TestRail-ready test cases in 8 minutes.
Free AI skill for Claude Code, Cursor, and any AI IDE. Uses Docling OCR + atomic Markdown generation + Excel export matched to your team's column structure.
Built because writing test cases manually is the most expensive thing QA engineers do.
Drop-in instruction sets that make your AI agent actually useful.
A curated catalog of reusable skills — each one a self-contained prompt you plug into Claude Code, Codex, or any compatible agent runtime and invoke by name. No configuration, no boilerplate.
- 📢 Telegram Channel: Join
- 💼 LinkedIn: Illia Karpenko
Open to: senior AQA roles, backend test automation consulting, and open-source QA tooling collaboration.
Tests should catch bugs, not create them. I keep that ratio in check.