Worked on the FZJ-IEK3-VSA/FINE repository to enhance the reliability of data integration by improving preprocessing for input matrices. Addressed issues with NaN values by automatically filling them with zeros and emitting targeted warnings, while also adapting validation checks for cases where diagonal elements are non-zero. This approach strengthened the robustness of transmission component checks and prevented silent failures in the esM integration pipeline. Leveraged Python and the Pandas library to implement these changes, focusing on data preprocessing and error handling. The work improved data quality signaling, reduced debugging time, and contributed to a more maintainable and reliable codebase.