Developed and integrated a vector-based knowledge retrieval system for the Tencent/WeKnora repository, focusing on enhancing search capabilities through Weaviate vector database integration. The work involved setting up a dedicated repository, configuring environment variables for secure deployment, and refactoring the tokenization method from a basic word approach to the more accurate 'gse' segmentation. Leveraging Go for backend development and database integration, the implementation emphasized scalable API development and robust text processing. This feature improved the retrieval engine’s ability to deliver richer, more relevant results, reflecting a methodical approach to performance, business value, and technical rigor within a short project timeframe.