At Klar, I own test automation strategy and coverage for the credit domain. I maintain a staging suite of about 750 tests that runs every 30 minutes, and reduced its runtime from about 45 minutes to 25–30 by regrouping tests across GitHub Actions runners.
I built an internal QA MCP server with a knowledge base of the microservice architecture and about 30 skills, integrated with Jira, Loki/Grafana, and ReportPortal. Compared with plain agents, it noticeably reduced token usage and sped up root-cause analysis.
I also built a Karate plugin for VS Code and Cursor for running and debugging tests, analyzing them, and scanning OpenAPI/Swagger coverage gaps. Developers and other teams use my credit-scenario simulation tooling to check cases without waiting for QA.
Earlier, I engineered automation frameworks and tests at BlockFi, Amusys Production, Glantus Ltd, Diebold Nixdorf BPO, and Polsoft Engineering. My Testpulse side project analyzes test quality with AST-based rules and security coverage checks against OpenAPI specs.

