Qingfeng Youchen contributed to both Eventual-Inc/Daft and volcengine/OpenViking, focusing on backend development and system optimization. For Daft, Qingfeng introduced a configurable minimum CPU parameter for Ray tasks, replacing a hardcoded value to enable dynamic resource allocation and improved parallelism, and enhanced build reliability by refining the Makefile clean process for Python-Rust integration. On OpenViking, Qingfeng refactored the resource ingestion pipeline, adding optional indexing and summarization while decoupling embedding metadata for scalability. These efforts, implemented in Python, Rust, and Makefile, addressed performance, maintainability, and deployment efficiency, demonstrating thoughtful engineering across distributed systems and data processing workflows.