Zendesk

Connect your ElevenLabs AI agents with Zendesk to manage support tickets, users, and organizations. This integration enables your agents to create and update tickets, search for existing records, manage users, and respond to incoming ticket comments.

This integration supports three authentication methods: the ElevenLabs OAuth app, your own OAuth client, and an API token.

Your Zendesk subdomain is the first part of your Zendesk URL (e.g., mycompany from mycompany.zendesk.com).

In the ElevenLabs integration setup, select the OAuth2 credential, enter your subdomain, and click Connect.

Zendesk shows an authorization page listing the access the ElevenLabs app requests: read access across your account, plus write access to tickets, users, webhooks, and triggers. Sign in as a user who holds the permissions your agent needs and click Allow. Zendesk redirects you back to ElevenLabs and the connection is created.

Authorize as an administrator if you plan to use Zendesk triggers — creating webhooks and editing triggers is restricted to Zendesk admins.

In Zendesk Admin Center, go to Apps and integrations > APIs > Zendesk API and enable Token Access.

Go to Apps and integrations > APIs > API tokens and click Add API token. Copy the token immediately — it is not shown again after closing the dialog.

Your Zendesk subdomain is the first part of your Zendesk URL (e.g., mycompany from mycompany.zendesk.com).

In the ElevenLabs integration setup, enter your email, API token, and subdomain.

In the ElevenLabs Zendesk integration setup, copy the redirect URL.

In Zendesk Admin Center, go to Apps and integrations > APIs > OAuth clients and click Add Client. Enter a name and a unique identifier.

Paste the redirect URL copied from ElevenLabs into the OAuth client configuration.

Save the OAuth client. Zendesk displays a secret — copy and store it securely. It is not shown again after closing the dialog.

In the ElevenLabs integration setup, enter the identifier as the client ID and provide the secret. Zendesk redirects you to authorize the connection.

Connecting with the OAuth2 credential grants ElevenLabs the following Zendesk OAuth scopes. Read access is granted account-wide, while write access is requested per resource so the token cannot modify resources the integration does not use.

A token never exceeds the permissions of the user who authorized it — the effective access is the intersection of these scopes and that user's Zendesk role.

These scopes do not apply to the other two authentication methods: a custom OAuth client requests read and write (full read and write access), and an API token carries all permissions of the Zendesk user it belongs to.

The table below lists every field the integration accesses in your Zendesk account, whether ElevenLabs stores it, and when and why it is accessed. Read means the value is used during processing but not persisted. Read & stored means it is persisted on the ElevenLabs conversation record. Written means the integration sends it to Zendesk.

The integration calls the Zendesk API only when a connected trigger fires or when an agent invokes a Zendesk tool. It does not poll your Zendesk account and does not export data in bulk.

Records returned by Zendesk tools are stored in the conversation transcript as part of the tool call result, so a tool that reads a record also stores it. Retention of stored data follows the privacy settings of the agent that handled the ticket, including transcript and PII deletion.

Add Zendesk tools to your agent to manage tickets, users, and organizations during conversations. Once the integration is connected, you can enable individual tools on your agent's configuration page.

The integration provides over 30 tools organized into the following categories:

• Tickets — create, update, delete, list, and search tickets. Bulk operations (create many, update many, delete many) are also available.
• Comments & tags — add public or internal comments to tickets, and add or remove tags.
• Users — look up, create, update, and delete users. Bulk create-or-update is supported for syncing user data.
• Organizations — retrieve organization details and list an organization's tickets.
• Search — run Zendesk search queries across tickets, users, and organizations (e.g., type:ticket status:open).

Creates a new support ticket. The agent collects details from the caller and opens a ticket on their behalf.

Searches Zendesk for tickets, users, or organizations using the Zendesk search query syntax.

Adds a public or internal comment to an existing ticket — useful for posting call summaries or follow-up notes.

Retrieves a user's profile so the agent can greet callers by name or verify their identity.

On your agent's configuration page, click Add tool and select Add integration tool.

Choose your Zendesk connection and toggle the tools you want the agent to use. You can enable as many or as few as needed — for example, a read-only triage agent might only need search and list tools, while a full-service agent might also create and update tickets.

Each tool's parameters are filled by the agent during the conversation based on what the caller says. You do not need to hard-code parameter values. However, you can optionally pre-fill or constrain specific parameters — for example, setting a default priority or group_id — to guide the agent's behavior.

The legacy integration uses three webhook tools to create the support agent. Review each tool's configuration in the tabs below.

Name: zendesk_get_ticket_comments Description: Retrieves the comments of a ticket. Method: GET URL: https://acmecorp.zendesk.com/api/v2/tickets/{ticket_id}/comments.json

Headers:

• Content-Type: application/json
• Authorization: (Secret: zendesk_key)

Path Parameters:

• ticket_id: Extract the value from the id field in the get_resolved_tickets results.

Tool JSON:

{
  "type": "webhook",
  "name": "zendesk_get_ticket_comments",
  "description": "Retrieves the comments of a ticket.",
  "api_schema": {
    "url": "https://acmecorp.zendesk.com/api/v2/tickets/{ticket_id}/comments.json",
    "method": "GET",
    "path_params_schema": [
      {
        "id": "ticket_id",
        "type": "string",
        "description": "Extract the value from the id field in the get_resolved_tickets results.",
        "dynamic_variable": "",
        "constant_value": "",
        "required": false,
        "value_type": "llm_prompt"
      }
    ],
    "query_params_schema": [],
    "request_body_schema": null,
    "request_headers": [
      {
        "type": "secret",
        "name": "Authorization",
        "secret_id": "zendesk_api_token"
      },
      {
        "type": "value",
        "name": "Content-Type",
        "value": "application/json"
      }
    ]
  },
  "response_timeout_secs": 20,
  "dynamic_variables": {
    "dynamic_variable_placeholders": {}
  }
}

Name: zendesk_get_resolved_tickets Description: Retrieves all resolved support tickets from Zendesk. Method: GET URL: https://acmecorp.zendesk.com/api/v2/search.json?query=type:ticket+status:solved

Headers:

• Content-Type: application/json
• Authorization: (Secret: zendesk_key)

Tool JSON:

{
  "type": "webhook",
  "name": "zendesk_get_resolved_tickets",
  "description": "Retrieves all resolved support tickets from Zendesk.",
  "api_schema": {
    "url": "https://acmecorp.zendesk.com/api/v2/search.json?query=type:ticket+status:solved",
    "method": "GET",
    "path_params_schema": [],
    "query_params_schema": [],
    "request_body_schema": null,
    "request_headers": [
      {
        "type": "secret",
        "name": "Authorization",
        "secret_id": "zendesk_api_token"
      },
      {
        "type": "value",
        "name": "Content-Type",
        "value": "application/json"
      }
    ]
  },
  "response_timeout_secs": 20,
  "dynamic_variables": {
    "dynamic_variable_placeholders": {}
  }
}

Name: zendesk_open_ticket Description: Opens a new support ticket. Method: POST URL: https://acmecorp.zendesk.com/api/v2/tickets.json

Headers:

• Content-Type: application/json
• Authorization: (Secret: zendesk_key)

Body Parameters:

• ticket: An object containing:
  • comment:
    • body: Detailed description of the support issue.
  • subject: A short subject line.
  • requester:
    • name: The full name of the requester.
    • email: A valid email address.

Tool JSON:

{
  "type": "webhook",
  "name": "zendesk_open_ticket",
  "description": "API endpoint to open a customer support ticket\nMake sure the authorization header is formated as \"Authorization: Basic <auth>\".",
  "api_schema": {
    "url": "https://acmecorp.zendesk.com/api/v2/tickets.json",
    "method": "POST",
    "path_params_schema": [],
    "query_params_schema": [],
    "request_body_schema": {
      "id": "body",
      "type": "object",
      "description": "Details for the support ticket",
      "required": false,
      "properties": [
        {
          "id": "ticket",
          "type": "object",
          "description": "This is the main ticket body which contains all of the information needed to open a ticket.",
          "required": true,
          "properties": [
            {
              "id": "comment",
              "type": "object",
              "description": "This is the comment with information about the issue.",
              "required": true,
              "properties": [
                {
                  "id": "body",
                  "type": "string",
                  "description": "Body of the issue. Include all relevant details for the issue. ",
                  "dynamic_variable": "",
                  "constant_value": "",
                  "required": true,
                  "value_type": "llm_prompt"
                }
              ]
            },
            {
              "id": "subject",
              "type": "string",
              "description": "Create a short subject line for the support issue. Add \"DEMO: \" before the subject.",
              "dynamic_variable": "",
              "constant_value": "",
              "required": true,
              "value_type": "llm_prompt"
            },
            {
              "id": "requester",
              "type": "object",
              "description": "The details of the support requester",
              "required": true,
              "properties": [
                {
                  "id": "email",
                  "type": "string",
                  "description": "The email address of the requester. This should look like \njohnsmith@hotmail.com\nYou MUST use the @ symbol and remove any spaces.",
                  "dynamic_variable": "",
                  "constant_value": "",
                  "required": true,
                  "value_type": "llm_prompt"
                },
                {
                  "id": "name",
                  "type": "string",
                  "description": "The full name of the requester. ",
                  "dynamic_variable": "",
                  "constant_value": "",
                  "required": true,
                  "value_type": "llm_prompt"
                }
              ]
            }
          ]
        }
      ]
    },
    "request_headers": [
      {
        "type": "secret",
        "name": "Authorization",
        "secret_id": "zendesk_api_token"
      },
      {
        "type": "value",
        "name": "Content-Type",
        "value": "application/json"
      }
    ]
  },
  "response_timeout_secs": 20,
  "dynamic_variables": {
    "dynamic_variable_placeholders": {}
  }
}

Configure Zendesk triggers to have your agent monitor and react to incoming ticket comments, providing first-line support.

In Zendesk Admin Center, go to Objects and rules > Business rules > Triggers and click Add trigger. Configure the conditions that determine which ticket events the agent should respond to (e.g., new tickets in a specific group, ticket comments with a certain tag). Note the trigger name — you will need it in the next step.

If you cannot save the trigger because an action is missing, add a simple action like adding an "agent is processing" tag to the ticket.

On your agent's configuration page, add a new trigger and select Zendesk Trigger. Configure the fields:

• Agent: the agent that handles incoming conversations.
• Trigger Rule Name: the name of the Zendesk trigger you created in the previous step.
• Daily Ticket Limit (optional): maximum number of tickets the agent handles per day. Leave empty for unlimited.

When you activate the trigger, ElevenLabs creates a webhook in your Zendesk account and adds it as an action to the trigger you specified. Deactivating the trigger removes the webhook and action.

In Zendesk Admin Center, go to Objects and rules > Business rules > Triggers and view your previously created trigger. You should see a new action added to it.

If you created another action previously, you can now remove it again.

Enable shadow mode on a Zendesk trigger to let the agent observe and draft responses without replying to customers directly. When shadow mode is active, the agent writes its responses as internal comments on the ticket instead of public replies. Only Zendesk agents and admins can see internal comments — the end user is not notified.

Shadow mode is useful for evaluating agent quality before going live. Review the internal comments alongside the actual support responses to compare accuracy and tone, then promote the agent to active mode once you are confident in its output.

When the agent responds to a ticket comment via the Zendesk API, that response is itself a new comment — which can re-trigger the agent and create an infinite loop. Add either of the following conditions to your Zendesk trigger to prevent this.

Exclude the service account from the trigger. Create a separate Zendesk user for the integration (e.g., ai-agent@yourcompany.com) and use this account's credentials when connecting in ElevenLabs. Then add this condition to your Zendesk trigger:

• Current User, Is Not, **\

**

Exclude API updates from the trigger. This filters out all updates made through the Zendesk API, regardless of which user made them:

• Ticket > Update Via, Is Not, Web Service (API)

• Zendesk API documentation
• Managing API tokens
• Zendesk triggers guide