Ze Hua Zou developed core data serialization and file I/O features for the apache/arrow and apache/iceberg-cpp repositories, focusing on Parquet and CSV support in C++ and CMake environments. He implemented Decimal32 and Decimal64 type handling in Parquet I/O, added configurable CSV header quoting, and introduced a Parquet writer for Iceberg-CPP with roundtrip integrity tests. His work included refactoring build systems, improving code clarity, and enhancing CI/CD reliability, particularly for macOS. By applying design patterns such as lazy initialization and updating file serialization logic, Ze Hua Zou improved startup performance, maintainability, and data correctness across both projects.