Platform development of a voice assistant : a LiveKit-based audio processing system
Xing, Haidong (2025)
2025
All rights reserved. This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:amk-2025121235721
https://urn.fi/URN:NBN:fi:amk-2025121235721
Tiivistelmä
The objective of this thesis was to design and implement a customizable voice assistant platform based on LiveKit, providing developers and enterprises with a flexible voice interaction solution. The platform would allow users to define the role, tone, and behavior of assistants via prompts and to generate embeddable pages for integration into web applications or interactive interfaces.
To achieve this goal, a scalable management and publishing system was developed, supporting speech recognition, real-time communication, speech synthesis, keyword-driven responses, and multi-assistant management. The front-end was implemented using React, Next.js, and TypeScript, while the back-end utilized Node.js and the LiveKit SDK, and Supabase PostgreSQL ensured data security and scalability. The entire application was hosted and continuously integrated by Vercel. Iterative development and user testing optimized system performance, configuration, and cross-platform embedding experience.
To achieve this goal, a scalable management and publishing system was developed, supporting speech recognition, real-time communication, speech synthesis, keyword-driven responses, and multi-assistant management. The front-end was implemented using React, Next.js, and TypeScript, while the back-end utilized Node.js and the LiveKit SDK, and Supabase PostgreSQL ensured data security and scalability. The entire application was hosted and continuously integrated by Vercel. Iterative development and user testing optimized system performance, configuration, and cross-platform embedding experience.