Worked on the spring-projects/spring-ai repository to enhance AI data fidelity and improve multi-turn interaction handling. Developed a feature in Java that enables the OpenAiChatModel to preserve additional response metadata, supporting richer downstream processing and analytics. Addressed a bug in the DeepSeek API integration by ensuring reasoning_content is included in tool-calling, which prevents HTTP 400 errors and maintains conversational context. The work involved backend development, API design, and comprehensive unit and integration testing. These contributions improved platform reliability and data quality, demonstrating a methodical approach to problem-solving and a focus on robust, test-driven engineering practices within the Java ecosystem.