During January 2025, Bao worked on the Kiln-AI/Kiln repository, focusing on backend development and data modeling to address data integrity issues in model data handling. He identified and fixed a Unicode encoding bug by explicitly enforcing UTF-8 encoding for all model data read and write operations, ensuring that non-ASCII characters, such as Vietnamese accents, are preserved without corruption. Using Python, Bao also implemented regression tests to validate the robustness of UTF-8 handling in data input and output. This work improved the reliability of multilingual content, supporting international business needs and reducing the risk of data loss in production environments.