During September 2025, Xjxlzlpzl developed automated robustness tests for TVF data loading in the apache/doris repository, focusing on improving reliability when handling abnormal data scenarios. They designed and implemented a comprehensive test suite using Groovy and SQL, targeting issues such as incorrect data types, null values in NOT NULL columns, formatting inconsistencies, and compression errors. By integrating these tests into the existing CI pipeline, Xjxlzlpzl enhanced automation and repeatability, enabling earlier detection of edge-case failures. Their work deepened regression coverage for TVF data loading, contributing to improved production stability and demonstrating strong skills in ETL, data validation, and test automation.