Built a knowledge-base Q&A system over PDF, DOCX, and Markdown documents using OpenAI APIs.

Design: Combined semantic chunking, embeddings, vector retrieval, and prompt design to reduce omissions and hallucinations in complex document Q&A, turning raw materials into an interactive knowledge base.

Designed a FedAvg-based multi-client training framework for privacy-sensitive behavior recognition, enabling model collaboration without moving local data.

Balance: Focused on the tradeoff between model performance, data locality, privacy protection, and deployability.

Contributed to a smart greenhouse digitization project, from county-level rollout planning to technical implementation design.

Delivery: Integrated data flows from two greenhouses, a compact weather station, and the ULAND IoT enablement platform, turning environmental monitoring and weather alerts into planting decision support.

Supported front-desk service and information-system maintenance at a university counseling center. The experience shaped my understanding that good systems should be efficient, privacy-aware, and humane.