Create and configure a Dialogflow agent that contains the text analyzer
artificial intelligence for the conversational channel.
When you create a
Dialogflow agent, the system generates a client access token and a developer access token
for the Dialogflow API. Record these tokens to use them for the Dialogflow text analyzer
configuration. When you create a Dialogflow agent, you also specify intents and entities
that the conversational channel uses to decide how to respond to a user.
-
In a web browser, log in to the Dialogflow developer portal with your developer
credentials.
-
Click Create new agent.
-
On the General tab for the agent, in the API
Version section, click V1 API.
Write down the unique client access token and developer access token that are
generated for the Dialogflow agent.
-
Click Save.
-
Click Intents, and then define the intents for your
conversational user channel.
- Optional:
In the Contexts section, add an input context.
Contexts record parameter values to pass them between intents.
-
In the Training phrases section, enter phrases
might trigger the intent when a user writes them.
-
In the Action and parameters section, enter action
and parameters.
Actions are sent to fulfillment after an intent is triggered.
Parameters are specific words or phrases that you are trying to collect
from users to complete a task.
-
Click Save.
-
Click Entities, and then define the entities for your
conversational channels.
You later map these entities with the case properties for your application.
You add synonyms for defined entities to a comma-separated list in the Unified Messaging, Web Chatbot, or Email channel
configuration.
-
Click Save.
What to do next: To use the Dialogflow component in Pega Platform for natural
language processing, configure the Dialogflow text analyzer in a Unified Messaging, Web Chatbot, or Email channel settings. Specify the
Dialogflow agent client access token and developer access token. For more information,
see Configuring the Dialogflow text analyzer for a channel.