Use the Flush Audio to flush any queued audio within a call flow. In Architect, the Play Audio action does not immediately play the audio file or text-to-speech (TTS). Instead, Architect queues the audio for later playback, and the flow execution continues. This behavior also occurs for other actions that include an audio input; for example, pre-transfer audio in a transfer action.

These operations and actions trigger an audio flush:

• Menus
• Flush Audio actions
• Collect Input actions
• Call Data and Call Secure Data actions
• Call Bot Flow actions and all third-party bot actions; for example Dialogflow, Lex, Nuance
• Detect Silence and Play Audio On Silence actions
• Hold Music action
• All Transfer actions
• Disconnect and Return to Agent actions
• Entering the first new Task or Menu that follows a Set Language action

For more information, see Task and state editor actions.

The Flush Audio action is available from the Play menu in the Architect task editor.