Hsuan-Lin (Nick) Huang

Scope: AI feature evaluation, company-wide AI tooling and E2E test automation

• Initiated a Claude Code marketplace and drove its company-wide adoption; with colleagues contributing, it grew into a company knowledge base of 40 plugins and 179 skills spanning engineering and business domains
• Architected an agent evaluation pipeline with promptfoo for production AI features, scoring accuracy, precision and F1 alongside latency and cost, and cut manual validation overhead by 70%
• Migrated E2E testing from legacy framework to TypeScript and Playwright, applying AI-assisted test generation and script refinement to reduce manual effort by 75% and implementation time by 80%

• Led a 6-member cross-functional team (2 Backend, 1 Frontend, 1 QA, 2 PMs), settling requirements and technical trade-offs directly with business teams while shipping a partner integration and two new business lines within a year
• Extended the existing platform architecture to carry two new business models, designing a forward-compatible data model that absorbed both without forking the order pipeline
• Stabilized Scrum delivery by institutionalizing task decomposition, estimation, and code review and documentation practices, cutting project delays by 30% and meeting overhead by 50%

• Executed a strategic database refactoring that consolidated 21 tables into 4 with zero downtime through dual-write migration, cutting development resources by 60%
• Cut image processing costs by 50% by re-architecting the pipeline onto AWS Lambda@Edge with CloudFront (Medium)
• Built and operated a production-grade monitoring stack on Prometheus, Grafana, and ELK with custom alerting, cutting error resolution time by 40%
• Cut crawler execution time by 60% by rewriting a Robot Framework crawler to fetch pages directly in Python, and containerized it with Docker to run in the cloud instead of on physical machines