Developed and enhanced text-to-speech capabilities for the pipecat-ai/pipecat repository, focusing on integrating Camb.ai’s MARS-8 TTS service with features such as voice selection, speed control, and interruption handling. Migrated the TTS stack to the official SDK, standardized model naming, and improved audio fidelity with model-specific sample rates. Expanded CLI functionality and maintained comprehensive documentation to support developer usability. Delivered a plugin for LiveKit agents, enabling multilingual speech synthesis. Employed Python, asynchronous programming, and API integration throughout, emphasizing reliability, maintainability, and test coverage. Addressed bugs and streamlined build automation, contributing to a robust, developer-friendly backend and plugin ecosystem.