Illia B.

I test API integrations and AI-agent tooling (MCP servers, REST APIs, agent gateways) and report what actually breaks: reproducible requests, evidence and a clear severity. Right now I do QA for a startup whose connectors let AI agents act in consumer apps over MCP and REST. The work covers unauthenticated sweeps, docs-vs-live drift checks, regression baselines and test-account re-login monitoring. What I do - API testing: Postman collections, auth and error contracts, schema and status-code checks, regression baselines - MCP / AI-agent tools: tool listings, auth flows, error envelopes, docs vs live behaviour - Python + pytest automation for repetitive checks; CI with GitHub Actions - Web and mobile (iOS/Android) functional testing; traffic debugging with Charles and Proxyman - SQL data validation (PostgreSQL, MySQL, MS SQL, DynamoDB) Background 5+ years in QA at Amdocs and Intellias (telecom, fintech, e-commerce), plus technical QA work during military service. Cut regression effort by 30% with Python scripts. How I work Short written updates. One bug per report: steps, expected vs actual, evidence. I don't present guesses as facts; a hypothesis is labelled as one. Tools: Postman, Swagger, Python, pytest, Charles, Proxyman, DevTools, Jira, TestRail, Git, Docker, Jenkins, JMeter.