---
sourceDocument: Australia Enable AI
sourceDocumentLink: https://www.servicenow.com/docs/r/australia/intelligent-experiences

 Release :

    - australia

ft:locale :

    - en-US

ft:publication_title :

    - Australia Enable AI

ft:clusterId :

    - platai

bundleId :

    - platai

workflow :

    - Platform


---

# Create a derived dataset

# Create a derived dataset {#ariaid-title1}

Release version: Australia  
Updated May 7, 2026  
![](https://www.servicenow.com/docs/portal-asset/ico-clock) 1 minute to read  
Create a smaller, derived dataset from an existing dataset using AI Data Kit. Use derived datasets to isolate specific records for focused ground truth labeling or evaluation without modifying the original dataset.

## Before you begin

Role required: sn_data_kit.admin

## Procedure

1. Navigate to AllNow Assist Data KitHome, and open an existing dataset in the Datasets tab.
2. Select the Create derived dataset button at the top of the Records list.
3. In the Add dataset info section, fill in the fields.  
   {#create-derived-dataset__table_xln_jnn_gdc__entry__2}

   | Field | Description |
   |-|-|
   | Dataset name | Name of the derived dataset. |
   | Dataset description | Description of the derived dataset. This field is pre-populated with a reference to the parent dataset. |
   | Tags | Keywords to help identify and search for the dataset. |
   [Table 1. Add dataset info fields]

   {#create-derived-dataset__table_xln_jnn_gdc}
4. In the Choose records section, select a selection method and choose the records to include.  
   A minimum of 10 records must be selected before you can create the dataset.
   * Manual Selection: Select the check box next to each record you want to include.
   * Sampling Selection: Select a sampling method from the drop-down menu, enter a number in the sample size box, and select Run to preview the sampled records.
5. Select Create dataset.  
   The derived dataset is added as a separate dataset.

## What to do next

After you have created the derived dataset, you can add ground truth to its records. For more information, see [Add a ground truth to each dataset record](https://www.servicenow.com/docs/2rdry7ykroMDzCguG1Duzw "Add a ground truth to each record in a dataset. A ground truth is an expected correct output for a given record. During evaluation in Now Assist Skill Kit, your custom skill's actual output is compared against the ground truth to measure accuracy and identify areas for improvement.").

*[\>]: and then


