Worked on the zipline-ai/chronon repository to enhance backend reliability by introducing compile-time validation for time-partitioned queries. Developed a mechanism in Python that enforces the presence of a partition column during the compile step, reducing runtime errors and improving developer feedback. Focused on data validation and unit testing, the work centralized configuration checks across group_by sources, join definitions, and bootstrap configurations. Comprehensive unit tests were implemented to ensure correctness and prevent regressions. This approach strengthened error handling by providing early, clear feedback on misconfigurations, ultimately improving the stability and maintainability of time-partitioned workloads within the core query paths.