Use an AI agent in a chatbot
EARLY ACCESSAdd AI agents to chatbots to enhance chatbot capabilities and reduce configuration complexity.
In your chatbot, use the Agent connector element to add an AI agent.
- Add the Agent connector element from the Apps from Exchange section in the Build side panel.
- Configure the element.
- Provide the input to the element.
- Save the response from the AI agent in an attribute.
- Either return the response to the end user or use it for further processing in your chatbot.
Configure the Agent connector element
Request
| Field | Description |
|---|---|
| Agent ID | Enter the ID of the AI agent. |
| User message | Specify the end user message that the chatbot received. Example: use the lastReceivedTextMessage predefined attribute. |
| Session ID | Specify the ID of the chatbot session. Example: use the sessionId predefined attribute. |
| Agent context (optional) | Select an attribute that contains additional context for the AI agent. |
Select Preview behaviour to preview if the response is a desired and reliable result.
Response
Save responses from the AI agent in attributes.
In the Response tab, complete the following fields:
| Field | Description |
|---|---|
| Save response code to attribute (optional) | Create an attribute to save the response code, then select it in this field. |
| Response body attributes | Select Add attribute. In the Attribute field, select the attribute from the list. In the Path field, enter the path to the attribute value in the JSON response hierarchy. |
Timeout
Specify how long you want to wait for the response. The minimum value is 1 second and the maximum value is 60 seconds.
Fallback
Configure the fallback to ensure the end user gets a response from the chatbot even if there is an error.
Use also for API failures
Enable the Use also for API failures option to trigger fallback in case of 4xx and 5xx errors.
