At MEXC, I tested core modules for the MOS activity landing-page system, including tasks, leaderboards, draws, games, and reward settlement. By investigating SQL and interface-chain bottlenecks, I helped reduce a task-data query from 10 minutes to under 3 seconds.
I also maintained interface and UI automation for MEXC’s activity platform, raising coverage of core flows above 90% and reducing regression testing from two hours to one. I used ChatGPT, Claude, and Cursor to generate test cases and scripts, while validating reward idempotency, inventory consistency, and fund safety under high concurrency.
At 成都同程旅行 and FITURE-成都拟合未, I tested web, H5, mini-program, and app products, including hotel booking, XR, AI drawing, and order-management systems. I also developed testing-efficiency tools and automated API checks, and tested invoicing, transaction reconciliation, and settlement workflows.

