Developed and integrated AI assistant performance metrics for the CherryHQ/cherry-studio repository, focusing on enhancing visibility into response latency and token generation speed. Leveraged React and TypeScript to update frontend components, enabling real-time display of these metrics within the user interface. Modified data structures to efficiently carry performance data from backend processes to the UI, supporting analytics and data-driven product decisions. The work provided actionable insights into AI response generation, allowing for targeted optimizations across the stack. Emphasized frontend development best practices and seamless user experience, delivering a feature that supports ongoing improvements in both product performance and user satisfaction.