Data Infra team Our team build a global, high throughput, low latency resource management and job execution service that enables user to run a variety of Analytics and ML applications spanning across regions and million machines • Onboard Container Test Library from scratch, aim to provide robust, low latency, container service to internal users • Container Test Library provide basic smoke test suite, and advanced stress test suite which include multiple CPU, RAM, I/O, Storage, and Throughput stress tests • Improved container test deployment efficiency by replacing time-consuming & manual overhead around creating and managing process; integrated with Meta’s internal infra service to be run as part of daily releasing test • Developed Unit Testing for library with more than 85% test coverage, ensuring a reliable engineering environment • Took ownership of library and launched the service within 10 weeks with detailed Wiki and multiple presentations

• Developed automation system for motherboard testing by Linux/UEFI tools • Overhauled a system by refining code and debugging test failure, leading to saving 25% of program runtime • Contributed to GitLab CI/CD workflows to improve test reliability and scalability, saving 5 hrs/week for QA team

• Predicted potential Dementia level by users’ daily walking path by self-developed Android App • Developed Machine Learning model(Mobilenet V1) for Dementia level prediction with 90% accuracy • Validated by physicians at NCKU Hospital, achieving 30% reduction in early diagnosis cost • Received Ministry of Science Technology(Taiwan) grant amount 20,000+ applications