Nikola D.

⭐ Top Rated · 100% Job Success I own release quality for teams shipping faster than they can test: Playwright suites that run in CI, and QA for AI-driven features where the output isn't deterministic. Clients hand me their repositories. I read the codebase, map how features actually behave, and file issues precise enough to act on without a round-trip for reproduction steps. One client put it this way: I'd soon know his projects better than he does. Clients usually come to me when one of these is true: - An AI feature shipped and nobody knows how to verify it - Regression takes too long and nobody trusts the suite - Tests are flaky, or test knowledge lives in one person's head - There's no QA process at all and someone has to build one What I've done: - Built a QA function from zero at a 20-year-old finance company with clients in 14 countries. Jira workflow, Xray, bug tracking conventions, and the onboarding guide they still use - Structured one Playwright (TypeScript) repo with 4 setups and 44 E2E cases for an enterprise client, so regression knowledge lives in the repo instead of in one engineer's head - Led a 3-person QA team on a client portal with Salesforce, SharePoint and Algolia integrations. Wrote the test plan at kick-off and wired Playwright runs into Azure DevOps with test-run reporting - Tested an InsurTech platform that turns claim data into AI-generated settlement recommendations. Scored output accuracy and completeness, and caught internal technical detail leaking into customer-facing results - 5+ years across FinTech, InsurTech, SaaS, e-commerce and enterprise products, as both QA Lead and hands-on engineer Core stack: - Automation: Playwright + TypeScript, Cypress, Selenium (Java), Appium + WebdriverIO - CI/CD: Azure DevOps pipelines, GitHub Actions, test-run reporting and release gates - API & data: REST (Postman, Playwright API), SQL validation, 3rd-party integration testing - Process: test strategy, risk-based prioritisation, regression suite ownership, QA process design, Scrum and Kanban - Also: WCAG 2.2 accessibility audits, cross-browser and responsive testing On AI features, I test measurable behaviour rather than generating more tests with an LLM: curated datasets, pass/fail rubrics scored against real cases, structured-output assertions, and checks for what should never reach the end user. I review automation pull requests, mentor other testers, and interview QA candidates. That's what owning quality looks like day to day, as opposed to executing test cases. Send me your stack and your main release bottleneck, and I'll propose a concrete first milestone.