Summary
Yung-hui Hsu is a Senior Data Scientist with seven years of experience building production-grade ML and AI systems, currently architecting distributed multimodal RAG pipelines and serverless AWS workflows at HP. He blends a strong statistical and computer-science foundation with hands-on expertise in edge CV (YOLO-Pose, DeepStream, TensorRT), self-supervised learning, and LLM customization (RLHF, PEFT, LangChain) to accelerate inference and improve retrieval performance. His work has delivered measurable wins—20x faster batch processing and a 1.5x Hit Rate@5 improvement—while his open-source DeepStream YOLO-Pose project earned community recognition. A former ecology researcher, he applies rigorous experimental design and explainable-AI techniques to make complex models interpretable for stakeholders. Colleagues rely on him for bridging research, engineering, and product needs, and for rapidly adopting new tools to solve practical problems.
7 years of coding experience
10 years of employment as a software developer
Master's degree, Ecology and Evolutionary Biology (IEEB), Best Oral Presentation Award for Master's Thesis, Master's degree, Ecology and Evolutionary Biology (IEEB), Best Oral Presentation Award for Master's Thesis at 國立臺灣大學
Big Data巨量資料分析就業養成班(BDSE), Big Data巨量資料分析就業養成班(BDSE) at Institute for Information Industry(資策會)
English, Chinese, Japanese