Voice Agent

• Speech-to-text (STT) - for understanding speech inputs
• LLM - for generating the agent text response
• Text-to-speech (TTS) - for generating agent speech audio

​

Run the Template

​

Start the Server

1. Download and extract the Inworld Templates and open it in an IDE.
2. Open a new terminal window.
3. Navigate to the voice agent server directory:
   Copy

   cd voice_agent/server
   

4. Make a copy of the .env-sample file:
   Copy

   cp .env-sample .env
   

5. Paste your Base64 Runtime API key into the .env file and set GRAPH_VISUALIZATION_ENABLED=true:
   .env

   Copy

   # Required, Inworld Runtime Base64 API key
   INWORLD_API_KEY=<your_api_key_here>
   
   # Optional, defaults to `false`
   GRAPH_VISUALIZATION_ENABLED=true
   

6. Install dependencies:
   Copy

   yarn install
   

7. Start the server:
   Copy

   yarn start
   

   The server will start on port 4000. You will see the path to your graph visualization printed in your terminal. Here is an example of the visualization:
   Graph visualization is currently not supported on Windows

​

Start the Client

1. Open a new terminal window.
2. Navigate to the voice agent client directory:
   Copy

   cd voice_agent/client
   

3. Install dependencies:
   Copy

   yarn install
   

4. Start the client:
   Copy

   yarn start
   

   The client will start on port 3000 (or the next available port if 3000 is in use) and should automatically open in your default browser.

​

Chat with Agent

1. Configure your agent in the UI:
   • Enter your name
   • Set the agent’s name
   • Provide a description for the agent
   • Define the agent’s motivation
   
2. Click “Start” to begin the conversation.
3. Chat with the agent:
   • Type text in the input field and hit Enter or click the send button
   • Click the microphone icon to use voice input
   • Click the copy icon to copy the conversation to the clipboard
   
4. Observe Telemetry
   • View dashboards, traces, and logs in the Inworld Portal

​

Next steps