Worked on the fuse-med-ml repository to enhance data ingestion flexibility by adding an option in OpReadDataframe that allows retention of the index column during dataframe loading. This feature, implemented using Python and YAML, improves data lineage and supports more robust downstream modeling pipelines by preserving index information throughout the data loading process. The approach involved modifying the data loading path, designing an option flag, and adhering to code linting and configuration management best practices. All changes were tracked through Git and issue management, ensuring reproducibility and alignment with project requirements while streamlining machine learning workflows for future development.