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

 Release :

    - brazil

ft:locale :

    - en-US

ft:publication_title :

    - Brazil Enable AI

ft:clusterId :

    - platai

bundleId :

    - platai

workflow :

    - Platform


---

# Add a dataset from an instance table

# Add a dataset from an instance table {#ariaid-title1}

Release version: Brazil  
Updated October 9, 2026  
![](https://www.servicenow.com/docs/portal-asset/ico-clock) 2 minutes to read  
Create a dataset from records in an instance table by choosing the columns to include and the conditions that select the records.

## Before you begin

Read access to the source table is required. The AI Data Kit roles don't grant it.

Role required: sn_data_kit.admin

## About this task

A dataset must contain at least 1 and no more than 10,000 records. Use the filter to keep the number of records within that range.

Dataset records keep the access controls of the source table. Anyone who views the records in the dataset must also be able to read the source table.

## Procedure

1. Navigate to AllAI Data KitHome.
2. In the An overview of your data assets section, on the Datasets tab, select Create dataset.
3. In the Choose data section, select Import data from Instance table.
4. In the Table field, select the table that contains the records that you want.
5. In the Columns field, select the columns to include in the dataset.
6. **Optional:** To add a column whose values are stored in another table, such as work notes, select Add column via scripting and fill in the fields.  
   {#add-dataset-instance-table__table_adit_scripted_column__entry__2}

   | Field | Description |
   |-|-|
   | Label | Display name of the column. Required. |
   | Name | Internal name of the column. Required. |
   | Type | Data type of the column values. The default is String. |
   | Script | Script that returns the column value for each record. The `current` parameter is the GlideRecord object for the record being processed. The editor includes a commented sample that adds work notes. |
   [Table 1. Add column via scripting fields]

   {#add-dataset-instance-table__table_adit_scripted_column}

   Scripted
   columns don't appear in the dataset preview.
7. **Optional:** To limit the records that are included, select Edit filter condition and build the filter conditions.  
   After you
   close the dialog box, the filter appears as an encoded query, such as `state=6`.
8. Review the records in the Preview dataset section.

   If the preview shows no records, check your filter conditions. Also confirm that you can read the source table, because the preview shows no records without read access.
9. Select Continue.
10. In the Add dataset info section, fill in the fields.  
    {#add-dataset-instance-table__table_adit_dataset_info__entry__2}

    | Field | Description |
    |-|-|
    | Dataset name | Name of the dataset. Required. |
    | Description | Description of the dataset that helps you and others find it later. |
    | Source type | Read-only. |
    | Tags | Keywords for the dataset. Press Enter after each tag. |
    [Table 2. Add dataset info fields]

    {#add-dataset-instance-table__table_adit_dataset_info}
11. In the Data governance section, select the I'm assuring to use data responsibly for AI Evaluation check box.  
    Add data isn't available until you select this check box.
12. Select Add data.  
    The dataset page opens with a state of In progress. Adding the records can take a few minutes. When it finishes, the state changes to Active.

    If a message says that a dataset must contain from 1 to 10,000 records, select Back and adjust the filter conditions.

## What to do next

After the dataset is Active, you can [scan it for sensitive data](https://www.servicenow.com/docs/qtH_BYLZAsGRafo7yaXM5A "Scan a dataset for personally identifiable information (PII) and other sensitive data, and cleanse the records that the scan finds."), [create a derived dataset](https://www.servicenow.com/docs/RzgYdXt7cK4IZriEAsSycg "Create a smaller dataset from an existing dataset by selecting records manually or by random sampling."), [add ground truth](https://www.servicenow.com/docs/3vCh~Vt0TiREOugFmDhZ5w "Add a ground truth, which is the real-world data that is used to train and test AI models to each dataset record. You can perform this task by using the AI Data Kit application."), or [add it to a data collection](https://www.servicenow.com/docs/lMBxQ5Pc_qQqWN5MglvFpQ "Combine one or more datasets into a data collection, choosing the columns and records to include. Then, publish the collection so that you can use it to evaluate a skill in AI Skill Kit.").

*[\>]: and then