Introduction

Edges are used to guide the caller through the script. They are created by drawing from one Action card to another.

When the voice agent will choose the edge to follow from among those coming out of the card, it will rely on the parameters of this edge.

To do so, 2 types of parameters will be taken into account:

• intent(s)
• condition(s) on context variable(s)

Edge’s conditions

Intent

When adding an intent, or a combination of intents, to an edge, the voice agent will wait for the caller to say something. If the sentence said by the caller matches an intent, then the voice agent follows the edge that contains this intent, otherwise, it can be directed to the Fallback script and will be treated as a NotUnderstood intent, which is a system intent.

Conditions using a single intent

You can add a single intent as a condition for the edge. A list of the available intents (created in the 3. Add intents and utterances page) will be shown. As you can add multiple intents to create a combination (we will see this use case later in the documentation), they will be treated as a tag in the input. That’s why you have to click on the cross icon on the right to actually validate the condition using the intent.

Conditions using a combination of intents

You can of course use a combination of intents, which makes this parameter very powerful. As an example, you may expect the caller to give both the last name and the first name at the same conversation turn. In this case, you can add several combinations, to understand the multiple ways that the caller can use to give both information:

• the first name and then the last name
• the last name and then the first name

Here, again, you will have to validate the full combination using the cross icon on the right. Then, you will be able to see the whole list of combinations of intents that can be understood at this conversation turn.

Condition on context variable