How do I configure voice and interaction settings for AI agents?

Last updated: November 5, 2026

Voice and interaction settings define how a voice AI agent sounds, responds, handles interruptions, and manages call behavior. Use these settings to tune the agent’s call experience for your workflow.

Before you start

  • Open the AI agent you want to configure.

  • Know the primary language and call style the agent should use.

  • Prepare any boosted keywords or pronunciation entries for important terms.

Steps

  1. Open the agent and go to the Editor tab.

  2. Scroll to the Voice and interaction section.

  3. Choose a language mode and primary language.

  4. Adjust speech-to-text mode and denoising.

  5. Select a voice.

  6. Configure voice speed, stability, and expressiveness.

  7. Set responsiveness and interruption sensitivity.

  8. Enable or disable backchanneling and choose its frequency.

  9. Configure reminders and pacing controls.

  10. Set maximum call duration and optional ambient sound.

  11. Add boosted keywords or pronunciation entries if needed.

  12. Choose a voicemail action and configure the voicemail message.

  13. Save your changes.

  14. Test the agent with preview or a controlled call.

Tips

  • Use single-language mode unless callers regularly switch languages.

  • Use accuracy mode for detailed calls or noisy environments.

  • Keep voice speed near a natural speaking pace.

  • Use low backchannel frequency for professional workflows.

  • Change one or two settings at a time so you can understand the impact.

Troubleshooting

  • The agent misunderstands callers: Switch to accuracy mode and enable denoising.

  • The voice sounds unnatural: Adjust speed or expressiveness in small increments.

  • The agent interrupts too often: Lower interruption sensitivity.

  • The agent responds too slowly: Increase responsiveness.

  • Voicemail behavior is wrong: Update the voicemail action and save again.