Voice

The Voice copilot template enables users to build a voice-enabled copilot that can provide an effective self-service, hands-free solution from a phone to reduce call volume and increase agent productivity. Customers can interact with a copilot by using copilot natural language and choosing the options (touch-tone) from a menu to resolve the issues faster.

You can create a new copilot in Copilot Studio. From the Home or Create page, under Start with a template, select the Voice template.

Screenshot of the Copilot Studio landing page, highlighting the Voice template option.

The primary building blocks of your voice-enabled copilot are constructed. Select Create to continue with the template.

To learn about regional settings, see Regional settings and data locations for organizations using Copilot Studio.

Prerequisite

  • Copilot Studio Account

Voice-enabled capabilities

When you select the Voice template, a copilot is created with several voice-related features included and enabled by default:

  • Telephony channel is enabled by default.
  • The Speech & DTMF (Dual Tone Multi Frequency) modality is enabled by default.
  • The predefined custom topic Main Menu.
  • An expanded version of the Conversation Start system topic that uses Main Menu.
  • Voice-related system topics include:
    • Silence detection
    • Speech unrecognized
    • Unknown dial pad press handles speech-related scenarios correctly
  • DTMF-enabled multiple choice options in the question node
  • The Test copilot chat mode is set to Speech & DTMF by default. This mode allows you to verify voice features, such as barge-in and DTMF, in your text input. You can see the speech response from your copilot in text form.

Limitations

AI-generated content can have mistakes, so don't forget to make sure it's accurate and appropriate. Review the Supplemental Terms.