Goal

Start a feedback dialog in a Conversational AI Cloud (CAIC) Voice project.

In a CAIC voice project, you can add a standardized feedback flow to your answers, which allows you to ask for feedback at specific points in the conversation, while still providing a consistent conversational user experience.

Product

Conversational AI Cloud

Prerequisites

The feedback flow requires the following content to be present in the project's culture:

• showFeedback addition, boolean value

• feedbackScore addition, integer value

• show_feedback event, should trigger a (feedback) dialog

Steps

The feedback flow will work as follows:

• The feedback flow is triggered whenever an answer contains the output addition showFeedback with a value of true 

• The show_feedback event will be triggered by Conversational AI Cloud (this is expected to trigger a dialog)

• The user steps through the dialog, where options can be presented to the end-user to determine their experience and feedback score

• When a dialog end node is reached:

  • The feedback flow is considered finished

  • The last input from the end-user in the dialog will be sent as the feedback label

  • The final node should also contain the feedbackScore out addition

  • The value of the feedbackScore addition is sent in as the score of the feedback

Potential Issues

If the feedback dialog flow is interrupted (because a different piece of content is triggered by the end-user):

• If one of the nodes in the dialog contained the feedbackScore value;  and, this node was selected by the end-user

• The last feedbackScore received will be sent in as the score of the feedback

• The last input from the end-user for the node that contained the feedbackScore will be sent in as the feedback label

Precedence of Events

When multiple events can be triggered only one event will be triggered because the user can only interact with one happening.

1. Livechat

2. Return event/Transaction End Event

3. Feedback

Good to Know

A few things to keep in mind when working with WhatsApp's standardized feedback flow:

• If the showFeedback addition is present in the middle of a dialog (i.e. not a dialog end node) or transactional dialog (i.e. not a transactional dialog end slot) then it will be ignored to maintain the conversational flow.

• If a return event is set on a transactional dialog node, or if the output addition livechat with a value of true is set on an answer, then the showFeedback addition will be ignored, as both livechat escalation or return events take precedence.

Background

You've already created a feedback dialog in Conversational AI Cloud.

The Conversational AI Cloud WhatsApp integration triggers a feedback event "show_feedback" in each Q&A and when you reach the end of a Dialog. You can leverage this to create a feedback flow using Dialogs. 

To disable feedback in an answer you use the output addition "showFeedback". This addition should be type "Yes/No" (boolean).
Set it to `false` if you want to disable the feedback.
By default all the feedback is disabled.
Define the feedback score using the Output Addition "feedbackScore".

​