This text provides a detailed account of creating a locally running voice assistant system, comprising a wake-word detection service, a voice assistant service, and a chat service. It also discusses the components and their interaction, as well as provides an example interaction with the voice assistant. The author highlights the surprising quality of the speech-to-text conversion and encourages others to try the project for various uses.
Build a Locally Running Voice Assistant
I was initially skeptical about Large Language Models (LLM) generating code snippets, but I was pleasantly surprised by the results. When my employer prohibited using online chatbot services, I built a locally running LLM service using open-source tools like HuggingFace and chainlit project. I also integrated voice interaction to enhance the user experience. This project will take you through the steps to build an assistant that allows vocal interaction with an open-source LLM, all running locally on your computer.
Architecture
The architecture involves three separate components: Wake-word detection service, Voice assistant service, and Chat service.
Chat Service
The chat service runs the open-source LLM called HuggingFace/zephyr-7b-alpha. It receives a prompt through a POST call, passes it through the LLM, and returns the output as the response.
Voice Assistant Service
The voice assistant service handles speech-to-text and text-to-speech conversions using locally running models. It allows for natural voice interaction and provides a seamless experience.
Wake-word Service
The wake-word detection service continually listens to the user’s microphone and triggers the voice assistant service when the wake-word is detected.
Example of an interaction with the assistant
Users can interact with the voice assistant using vocal commands, receiving spoken responses to their queries.
If you want to evolve your company with AI, stay competitive, and use AI to your advantage, consider how a locally running voice assistant can redefine your way of work.
For AI KPI management advice and continuous insights into leveraging AI, connect with us at hello@itinai.com or stay tuned on our Telegram or Twitter.
Spotlight on a Practical AI Solution
Discover practical AI solutions like the AI Sales Bot from itinai.com/aisalesbot, designed to automate customer engagement and manage interactions across all customer journey stages.
Explore how AI can redefine your sales processes and customer engagement at itinai.com.
