Adatasetis an organized collection of information that analyst copilot uses to create reports. Several prebuilt datasets are included, organized by business process that you can use to ask questions of your business information. However, you can also combine information from multiple datasets into a new, custom dataset using dataset builder.

What's my plan?
All Suites Professional, Enterprise, or Enterprise Plus
Support with Explore Professional or Enterprise

Summary: ◀▼

Use analyst copilot datasets to build reports from prebuilt data like tickets, agents, and IT assets, or combine sources into a custom dataset with dataset builder. You can join objects, choose columns, and create datasets for reporting without needing database expertise. Datasets refresh once a day, and new datasets may be added through the EAP.

This feature is currently in an Early Access Program (EAP). You can sign up for the EAP here.

A dataset is an organized collection of information that analyst copilot uses to create reports. Several prebuilt datasets are included, organized by business process that you can use to ask questions of your business information. However, you can also combine information from multiple datasets into a new, custom dataset using dataset builder.

Watch the following video to see dataset builder in action:

Using dataset builder (2:27)

This article contains the following topics:

  • Understanding the prebuilt datasets
  • Creating custom datasets
  • Adding descriptions to your custom datasets

Understanding the prebuilt datasets

The agentic analytics EAP includes the following prebuilt datasets:

  • Tickets: Gain insights into ticket volume and activity. Things like ticket status, groups, resolutions, reply times, satisfaction score, as well as AI suggestions, enhanced writing and intelligent triage Copilot features. Data is captured at the ticket level in the current state.
  • Agents: Get insights into your support team members, including profile details, role, skills, and group and brand memberships. Data is captured at the user level in the current state.
  • IT assets: Get insights into your IT asset inventory, including asset details, ownership, lifecycle status, and assignment history. Data is captured at the asset level in the current state.
  • AI agents: Gain insights into your AI agent performance. AI agents consists of three datasets:
    • AI agent conversations
    • AI agent conversation knowledge
    • AI agent conversation use cases

When you create a report, you can select the dataset you want to use or you can create your own.

Follow the Agentic Analytics EAP page to discover when new datasets are released.

Creating custom datasets

While you can't use multiple datasets in your report, you can combine information from multiple datasets into a new dataset known as a custom dataset. The dataset builder uses a graphical interface, so minimal knowledge of databases is necessary.

To create a custom dataset

  1. In Analytics, click the Home () icon in the sidebar.
  2. Click the create icon (+), then select Dataset.

    Dataset builder opens.

  3. Click Add data source.

    Dataset builder adds a standard object. This doesn't contain any information until you configure it.

  4. Click the standard object you added to open its options.

  5. In the object options, configure the following as needed:
    • Type: Change the object type from the default of standard. Choose from standard, custom, and prebuilt objects, or custom fields.
    • Object: Choose the object from which you want to choose data; ticket, user, or organization.
    • Select columns: After you select the type and object, the extracted columns will be displayed. Select the columns you want to include in your custom dataset.
  6. If you want to add data from other sources, double-click in a blank area of dataset builder to display the dataset menu.

    Tip: To delete an object, click to select it, then press Delete.
  7. From the dataset menu, choose one of the following options:
    • Extract Zendesk data: Adds a new object to the dataset builder. You can configure this in the same way you configured the initial object. The available object types are:
      • Standard object: Includes tickets, users, and organizations.
      • Custom object: User-defined data to track business-specific information (such as assets, contracts, or inventory) beyond standard objects such as tickets or users.
      • Prebuilt object: Prebuilt Objects (Tickets, Agents Daily Snapshot, Agents) - these are curated for you by Zendesk
        • Tickets: Gain insights into ticket volume and activity including ticket status, groups, resolutions, reply times, satisfaction score, as well as AI suggestions, enhanced writing, and intelligent triage copilot features. Data is captured at the ticket level in the current state.
        • Agents daily snapshot: Analyze agent activity at the end of any given date. Examples include Copilot features usage, assigned tickets, and solved tickets. Data is recorded at the agent level and captured daily.
        • Agents: Get insights into your support team members, including profile details, role, skills, and group and brand memberships. Data is captured at the user level in the current state.
      • Custom fields: Includes tickets, users, and organizations.
    • Join: While you can add multiple data sources to your custom dataset, they will have no relationship to each other unless you join them. To create a join, drag from the L or R nodes in the join to the objects you want to join. The following joint types are available:
      • Left join: Returns all data from the first object and matches from the second object.
      • Right join: Returns all data from the second object and matches from the first object.
      • Inner join: Returns only data that exists in both objects.
      • Outer join: Returns all data from both objects.
    • Prefix: Required to make sure that if you join two data sources together that have a column with the same name such as "ID", then the joined column will have a prefix such as "claims_ID" if joining to a claims custom object.
    • Drop: Add a Drop object if you want to remove columns from a connected object.
    • Register dataset: Adds a Register dataset object. Connect this to your data source object. Then
  8. Click the Save as panel, enter a Name, Description, and Project for the new dataset, then click Save.

Your new dataset is scheduled to run. After it's complete, you can use it to create reports.

Note: The content of datasets update once a day at midnight in the account's timezone.

Adding descriptions to your custom datasets

When you add information to a report using report builder, you can hover over any dimension or measure to see a description of it.

When you create a custom dataset, you can specify the description that will be displayed in report builder. You can specify descriptors manually, or analyst copilot can generate them for you. Adding descriptors makes it easier for report builders to understand your dataset and also improves the performance of memories and suggestions.

To specify custom dataset descriptions

  1. On the analyst copilot home page, hover over your custom dataset, click the options menu, then select **Edit descriptors**.

  2. On the dataset descriptors page, perform one of the following actions for each of your dataset columns:
    • Enter a new description for the column.
    • Enter a few keywords, then click Prep. Analyst copilot automatically generates the description based on your keywords.
    • Leave the custom description blank and click Prep. Analyst copilot automatically generates the description.

  3. When you've finished specifying descriptors, click Save.

When you build a report using your custom dataset, your new descriptors will be displayed whenever you hover over a dimension or measure.

Powered by Zendesk