AI Agent Fields

Overview

AI Agent Fields can be used to temporarily store information during an agent interaction with a customer that can be used to make decisions within your flow.

AI Agent Fields are similar to Custom Fields, but the information is not permanently stored with the conversation, it’s only used to keep track of data during the agent session.

As an example, you may want to increment a counter each time the customer lands on a fallback behavior. If they have been there more than 3 times, you could offer them the ability to speak to an agent.

Creating AI Agent Fields

Agent Fields can be created or edited from within the Configuration Panel by selecting the Data Fields option:

You must configure a Label, Description, ID and Type for each field. The type of the fields can be Text, Boolean, Integer or Select.

Using Agent Fields in Analytics

As mentioned above, agent fields are generally not available outside of that particular agent, however you can enable up to 10 agent fields to be leveraged in Analytics by using the Include in Analytics checkbox at the bottom of the form:

🚧

As you might expect, Agent Fields are only available in analytics for events in the Bot category:

Additionally, you'll need to ensure you've selected the appropriate User before events will show up:

Using Agent Fields

Once an Agent Field is created, it can be used in Decisions and in Set Field and Add to Field (if it’s an Integer) Actions in your flow.

Agent Fields can also be created when adding a Set Field or Add to Field action, or a Decision Condition. Selecting the Add New Agent Field button next to the field selector drop-down will open the dialog to create a new Bot Field.

Upon choosing this action, you must then choose the field that you would like to set and provide a corresponding value:

Adding to a Numeric Field

Your Agent can also be configured to add to numeric fields. Simply choose the ‘Add to a numeric field’ action and specify the amount and which field it should be added to.

In this case, testNumeric’s value will be increased by 5. Subtraction can be achieved by specifying a negative amount. This action is useful when you want your bot to maintain some sort of running total or score. For example, you might maintain a score based on the customer’s responses to various questions and ultimately use the score to decide how to route the conversation.

If you don’t see any available fields to add to then you need to request that a numeric custom field be added to your site. Please send your request to [email protected].

🚧

Custom Field Vs Agent Field

If you want information to be available as part of the conversation details to be reviewed later, like in a Closed Conversation CSV export, you should use a Custom Field to store the information, not an Agent Field.

Agent Fields are stored within the Agent definition, and created right in AI Studio while configuring your flow. This means that if you export an Agent with a set of Agent Fields in one tenant, you can easily import it into a different tenant without having issues with dependencies that may occur with custom fields.