Worked on backend reliability and data validation for open source data infrastructure, focusing on bug fixes in Apache Iceberg Python and TrinoDB/Trino repositories. In Apache Iceberg Python, addressed robustness in the SetStatisticsUpdate path by refining validation logic to support both Pydantic model instances and dictionaries, preventing runtime errors and ensuring stable statistics updates in production pipelines. In TrinoDB/Trino, resolved a regression in SQL domain translation by correcting type checks for the starts_with function, restoring expected query behavior. Employed Java, Python, and SQL, with an emphasis on backend development, testing, and maintaining compatibility while improving system correctness and reliability.