Worked on enhancing the DoclingParser component within the Shubhamsaboo/RAG-Anything repository, focusing on improving the robustness of document parsing for complex, nested structures. Applied Python programming and data parsing expertise to refactor the parsing logic, ensuring accurate page index calculations and reducing errors in downstream processing. The changes were validated through targeted commits, including a core fix that addressed reliability issues in content extraction. This work improved the maintainability and readability of the codebase, laying a stronger foundation for future enhancements and supporting more reliable retrieval workflows in document processing applications that rely on precise structure handling.