Developed and integrated the SLNG plugin for Speech-to-Text and Text-to-Speech within the livekit/agents repository, enabling multi-voice AI capabilities for real-time agent workflows. The work focused on plugin-based architecture, allowing scalable audio processing and extensible AI integration across the platform. Leveraging Python and WebSocket programming, the solution facilitated live transcription and synthesis, supporting dynamic voice interactions in agent environments. The implementation established a foundation for future SLNG-based extensions, emphasizing maintainable code and collaborative development practices. No major bugs were addressed during this period, with efforts concentrated on delivering robust feature functionality and seamless integration into existing LiveKit Agents infrastructure.