Set up user input for a voice survey flow

The User Input menu lets you set up the voice survey flow behavior when the voice survey flow cannot confirm a caller’s match or input. Here, you can also set up barge-in settings, confirmation rejection prompts, and confidence and confirmation threshold settings. Use built-in prompts or build logic around replies that the survey participant hears when the voice survey flow does not detect a match or inputs, or encounters a confirmation error.

Under Settings, click User Input. 

Name Description

General

Configure the following settings:

  • Voice Barge-In: Enable or disable whether a survey participant can interrupt voice survey audio. This global setting allows participants to bypass listening to a question. For example, when your survey participant does not want to wait for the entire question. The global barge-in feature is available at points in the conversation when the voice survey flow expects an answer from the survey participant; for example, in an Ask Survey Question action. A participant cannot interrupt at points in the conversation when the voice survey flow does not expect an answer, such as in a Communicate action. 

Note: When you enable this setting, you enable it for all “Ask” actions in the voice survey flow that you are currently editing.
No Match Detected

Configure the following settings:

  • Maximum Number of No Match Retries: Set the maximum number of times that the voice survey flow attempts to match the user input. If the voice survey flow exceeds the maximum number of retries, then the event handling Recognition Failure Event settings trigger.
    Note: For Ask Survey Question actions, if the maximum number of no max retries is exceeded, Architect skips the current Ask Survey Question action and the flow moves onto the next action in the flow.
  • No Match Apology: The apology communication that prefixes the current Ask action’s No Match communication when the voice survey flow receives user input that was not a valid match.
No Input Detected

Configure the following settings:

  • Maximum Number of No Input Retries: Set the maximum number of times that the voice survey flow retries if it does not receive input from the user. If the voice survey flow exceeds the maximum number of retries, the following rules apply:
    1. If enabled on the action, then the flow takes the Max No Inputs path. You cannot enable the Max No Inputs path on the Ask for Slot action and the Ask for Yes/No action in a voice survey flow.
    2. Architect skips the current Ask Survey Question action and the flow moves onto the next action in the flow.
  • No Input Apology: The apology communication that prefixes the current Ask action’s No Input communication when the voice survey flow does not receive input from the user after the relevant timeout.
  • Voice No Input Timeout: Set the maximum time to wait for the survey participant to respond. The default setting is 7 seconds.
Confirmation Errors

Configure the following settings:

  • Maximum number of Confirmation Rejections: Set the maximum number of times that the voice survey flow allows a survey participant to reply with a “no” response when the voice survey flow attempts to confirm an input. If the voice survey flow exceeds the maximum number of retries, then the event handling Recognition Failure Event settings trigger.
  • Confirmation Rejection Apology: The apology communication that prefixes the current Ask action’s No Match communication when the survey participant replies with a “no” after the voice survey flow attempts to confirm the participant’s last input.
  • No Match Confirmation Apology: The apology communication that prefixes the current confirmation question when the voice survey flow receives input from the survey participant that does not match “yes” or “no.”
  • No Input Confirmation Apology: The apology communication that prefixes the current confirmation question when the voice survey flow does not receive input from the survey participant after the relevant time out during confirmation.
Speech Recognition Parameters

Configure the following settings:

  • Speech Detection Sensitivity: Set the sensitivity of speech detection to configure how to treat noise (for example, breathing) and when to trigger speech detection. A value of 0 configures the system to be the least sensitive to noise while a value of 1 makes the system highly sensitive to noise and triggers speech detection on any noise. When you use a NOT_SET value, the default value of 0.5 is used.
    • Minimum value: 0
    • Maximum value: 1
    • Default value: 0.5
  • Max Speech Timeout: Set the maximum duration of an utterance collected from the user. A duration value less than or equal to 0 milliseconds means no timeout. When you use a NOT_SET value, the default value is used. The default value is 5 minutes for Google Cloud STT, 10 minutes for Microsoft Azure Cognitive Services STT, and 22 seconds for Nuance Recognizer as a Service.
    • Minimum value: 0 milliseconds
    • Maximum value: 30 seconds
DTMF Settings

Configure the following settings:

  • Inter-Digit Timeout: Sets the length of time to wait, in seconds, from when the survey participant presses a digit until the participant presses the next digit. The default inter-digit timeout is 3 seconds. For example, a survey participant begins to enter a four-digit number and is distracted after pressing the first digit. After 3 seconds, the system prompts the survey participant to start data entry over.
    Click + and – to set the number of seconds the system waits between each entry before timing out.
  • Terminating Character: Defines the character that the survey participant can use to indicate that data entry has been finished. “*” and “#” are the allowed terminating characters.