Williams Liu

Williams Liu, based in Taipei.

10+ years of software testing discipline — from gaming (lead QA on MapleStory; later managing a 50-person QA team) through education, media, blockchain, and AI platforms. Black-box testing and requirements verification across web, app, and API are the foundation. I work from the user's vantage combined with spec logic — boundary conditions, exception flows, special-scenario simulations to surface latent risk.

Current focus: bringing QA rigor into the design and polishing of AI agents.

I've hand-built a dual-layer QA Orchestrator architecture: each layer holds its own test memory and workflow playbook, with ironclad governance — release-gate confirmation, self-learning from defects, sensitive-data exclusion from version control, Prod-environment guardrails — all Git-versioned and reproducible. With AI cowork I've wired the full test loop: from Jira ticket analysis, to test plan and case authoring, mixed manual + automated execution (Playwright / Midscene), writeback to cases and tickets, and final test report generation.

Tooling: TCM on Jira / MeterSphere / Qase; automation on pytest / Playwright + Cursor; security cases via OWASP (authz, SQLi, XSS); push notifications through Slack incoming webhook (QA Weekly Bot). I've also led a 50-person rotating-shift QA team — staffing interviews, KPI design, cross-team coordination, on-site vendor exchange in the Philippines.

I look for flat, agile, goal-oriented teams that value real output and a collaborative spirit.