Getting started with an agentic AI agent (EAP)

What's my plan?

Add-on

AI agents - Advanced

Agentic AI agents are currently in an early access program (EAP). You can sign up for the EAP here.

Agentic AI agents are designed to solve complex user requests with limited direct human supervision, and without requiring scripted conversation flows. These AI agents increase your automation rates and reduce the load on human agents and admins, while still providing the level of service your customers expect.

This article contains the following topics:

• Signing up for the EAP
• Getting started with your agentic AI agent

Related articles:

• About agentic AI agents (EAP)
• Creating generative procedures for agentic AI agents (EAP)

Signing up for the EAP

Agentic AI agents are currently available only as part of an early access program (EAP). To be considered for participation in the EAP, fill out this form. Zendesk will carefully review each submission to ensure the best fit for the program.

If you’re accepted into the EAP, Zendesk will create an agentic AI agent for you and get back to you as soon as possible.

After the agentic AI agent is created for you, you can get started by performing the tasks below.

Getting started with your agentic AI agent

After your agentic AI agent is created for you, there are several mandatory and optional setup tasks:

• Configure an AI agent persona
• Import a knowledge source
• Create use cases
• Create generative procedures or dialogues
• Create actions, content entities, and API integrations (Optional)
• Continue configuring the AI agent settings (Optional)

Configure an AI agent persona

The first step in your agentic AI agent setup is configuring an AI agent persona that controls the identity of the AI agent and how it presents itself to your customers. A persona gives the AI agent the context of who it’s working for, what your company does, what your products are, and so on. If you don’t configure an AI agent persona, a technical error will appear for every message the AI agent receives.

For instructions, see Customizing the persona and tone of voice for your advanced AI agent.

Import a knowledge source

For an agentic AI agent to function with generative procedures, it must have at least one knowledge source imported. The knowledge source can be a help center, CSV file, or web crawler.

Note: At the moment, generative procedures used by agentic AI agents don’t work with search rules for knowledge sources.

For instructions, see Importing knowledge sources for an advanced AI agent.

Create use cases

Next, create use cases that allow the AI agent to understand customer requests and follow the correct generative procedures or dialogues. If you don’t create any use cases, the AI agent forms responses using only the content of your imported knowledge source, not procedures or dialogues (see the next section).

For instructions, see Creating use cases for advanced AI agents.

Create generative procedures or dialogues

The most powerful step in agentic AI agent setup is to create generative procedures. These procedures should reflect your business policies so that the AI agent can resolve customer requests in line with your policies.

For instructions, see Creating generative procedures for agentic AI agents (EAP).

In addition to generative procedures, you can also create dialogues that allow the AI agent to respond to customer requests according to scripted conversation flows.

For instructions on creating dialogues, see Using the dialogue builder to create conversation flows for advanced AI agents.

Agentic AI agents that use both generative procedures and dialogues are sometimes referred to as hybrid AI agents. You can decide which use cases should use the more flexible generative procedures, and which should use more prescriptive dialogues.

For instructions on choosing between dialogues and procedures, see Configure whether a use case uses a dialogue or generative procedure (EAP).

Create actions, content entities, and API integrations (Optional)

To leverage helpful details in either generative procedures or dialogues, you can create:

• Actions that allow the AI agent to perform actions based on the details of the session or your customer relationship management (CRM) system. For instructions, see Creating and adding actions for advanced AI agents.
• Content entities that hold pieces of information in visitor messages that have specific meaning, such as the user’s email address. For instructions, see Creating content entities in conversation flows for advanced AI agents.
• API integrations that leverage information from other backend systems you use during your workflows. For more information on using the integration builder to create API integrations, see About the integration builder for advanced AI agents.
  Note: For APIs to be used by an agentic AI agent at the right time and in the right context, it’s extremely important to add a clear description to your APIs while building them in the integration builder.

If your AI agent was converted from zero-training to agentic AI, and you already had existing actions, content entities, or API integrations for that AI agent, you can use those in generative procedures or dialogues without recreating them.

Continue configuring the AI agent settings (Optional)

At any point, you can always continue configuring the AI agent’s settings. For details, see About the settings for advanced AI agents.