Skip to main content
The Audio section, found under Voice in Agent Studio, gives you access to every audio clip your agent uses on calls. From here you can preview and regenerate, individual clips — useful for fine-tuning specific lines that the agent may say awkwardly (i.e. awkward cadence, speed, etc.)

​
How to find and adjust audio

  1. In the PolyAI platform and go to Voice in the left-hand navigation.
  2. Click the Audio tab to see the full list of generated audio clips.
  3. Use the Add Filter button and use the filter “Transcript contains” if you are looking for a specific clip Screenshot 2026 07 16 At 3 59 07 PM
  4. To hear what the clip currently sounds like, click the the Play button on the left
  5. If you want to regenerate the clip, click the three-dot menu at the far-right and click Edit
Screenshot 2026 07 16 At 4 04 46 PM
  1. Click the Regenerate button on the left to regenerate the audio clip. Once it’s regenerated, it will automatically play.
Screenshot 2026 07 16 At 4 05 48 PM
  1. You can click the Regenerate button until the clip sounds the way you want it to. Once you are happy with it, click Save, then Proceed.
  2. The audio clip in the list will now be tagged with “Regenerated” and the agent will use the regenerated audio going forward whenever it says that phrase.
Screenshot 2026 07 16 At 4 10 34 PM