Worked on the eosphoros-ai/DB-GPT repository to deliver per-document tracking for document loading, enhancing document management and traceability within the system. Implemented this feature by introducing an optional file_id parameter to the aload_document method in IndexStoreBase, allowing for more granular tracking of individual documents. Addressed an argument mismatch in a knowledge graph function call to improve reliability and prevent runtime errors. Utilized Python for backend and API development, focusing on data governance and retrieval accuracy. The work contributed to improved system robustness and maintainability, reflecting a methodical approach to backend feature development and technical problem-solving.