Worked on the lancedb/lance repository to implement row-level identifiers in take results, enhancing data traceability and correctness for Lance datasets. Developed logic to fetch and project _rowid and _rowaddr columns within the schema, ensuring these identifiers are consistently returned in record batches. Expanded the schema projection mechanism to automatically include new identifiers across various configurations, supporting robust data engineering workflows. Added comprehensive cross-configuration tests to validate the presence and reliability of row identifiers. Utilized Python and Rust for backend development and testing, focusing on improving data integrity and schema management without introducing new bugs during the development period.