At SAR Group, I own quality and release go/no-go decisions across five live B2B and B2C products. I introduced developer-side and CI quality gates, and built web and mobile automation alongside API, performance, and security testing.
I also build and evaluate AI-assisted QA tools, including LLM agent evaluation harnesses with golden datasets, deterministic scoring, and regression baselines. My open-source work includes Verdict, a CI failure triage engine, and tools for browser-agent and LLM security evaluation.
