During September 2025, Zheng Niu contributed to the Azure-Samples/cognitive-services-speech-sdk repository by developing and integrating custom Text Normalization sample data. He created new sample data files and implemented formatting changes using C++ to support testing and demonstration of custom Text Normalization workflows. Zheng focused on sample data generation and text normalization techniques, enabling users to more easily onboard and validate their own text normalization pipelines within the SDK. While the work was limited in scope to a single feature, it addressed a specific need for practical examples, improving the usability of the SDK’s custom Text Normalization capabilities for developers.