During January 2026, work focused on enhancing PPTX document parsing within the DS4SD/docling repository. Using Python and backend development skills, a new feature was implemented to improve the extraction and grouping of bullet points under their correct subheadings, addressing previous inconsistencies in document structure. This parsing enhancement increased the reliability and readability of exported PPTX content, reducing the need for manual cleanup and supporting more accurate downstream analytics. The approach centered on robust data parsing techniques, ensuring that the document’s logical hierarchy was preserved throughout the workflow and enabling smoother integration with subsequent documentation and processing pipelines.