Developed a foundational Text-to-Speech capability for the livekit/agents repository, enabling audio synthesis from text input by integrating the Speechmatics API. Focused on backend development using Python, the work introduced asynchronous programming patterns and robust data handling to support customizable voice options, allowing bots and assistants to adopt multiple personas for diverse use cases. The implementation emphasized minimal surface area changes and maintained backward compatibility, ensuring seamless adoption within existing systems. Delivered as a cohesive feature with clear documentation and commit practices, this addition improved user engagement and accessibility while laying groundwork for future automation and conversational enhancements in the platform.