Speech and chat settings overview (Customer Conversation) (preview)

[This article is prerelease documentation and is subject to change.]

Every agent you create with the Customer Conversation harness supports both Speech and Chat conversations by default. You don't choose a channel modality when you create the agent. Instead, you configure how the agent should behave for each modality within the same agent. Voice is a configurable feature of the Speech modality, covering settings such as voice font, DTMF behavior, speech sensitivity, silence detection, and latency messaging.

Modality-specific instructions

In the Instructions section of your primary agent or a sub-agent, use the Speech and Chat tabs to add guidance that applies specifically to Speech conversations or specifically to Chat conversations, in addition to instructions that apply to both. Use this to account for differences between the two modalities, such as:

  • Speech responses need to be concise and easy to follow when heard rather than read.
  • Speech channels don't support end-user sign-in, so instructions and knowledge sources that require end-user credentials should only apply to Chat conversations.
  • Chat conversations can include rich content, such as cards, links, and images, that isn't meaningful over a Speech channel.

Voice authentication

If your organization enables biometric authentication and fraud detection capabilities in the Microsoft Contact Center Intelligence portal, every agent is pre-enabled with a Voice Authentication setting. Voice Authentication works together with the Biometric Authentication capability to identify and verify callers during a Speech conversation. For steps to confirm this setting, see Configure agent details and security (Customer Conversation). For voice-specific settings, see Configure voice settings (Customer Conversation).

  • Last updated on