Worked on integrating Neuphonic Text-to-Speech into the pipecat repository, delivering end-to-end streaming audio capabilities with multilingual support through both WebSocket and HTTP. Leveraged Python and JavaScript to refactor audio processing, streamline model parameter handling, and align environment configurations. Expanded internationalization by updating documentation in pipecat-ai/docs to include French, Portuguese, Russian, Hindi, and Chinese language codes. In the TEN-framework/ten_framework repository, implemented a real-time TTS extension using Server-Sent Events for dynamic audio streaming, while also improving codebase hygiene. Focused on backend development, API integration, and asynchronous programming to enhance maintainability and enable real-time, multilingual conversational AI features.