> This page is for Docs.

> For clean Markdown of any page, append .md to the page URL.
> For a complete documentation index, see https://zuora.ferndocs.com/llms.txt.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://zuora.ferndocs.com/_mcp/server.

# Create an AI-assisted Data Loader job

> Learn how to create an AI-assisted Data Loader job, map source fields, resolve validation issues, and submit the job.

Learn how to create an AI-assisted Data Loader job, map source fields, resolve validation issues, and submit the job.

* Prepare your source file using a supported format: CSV, Excel, JSON, or JSON Lines.
* Organize your data for import.
* For update operations, identify the fields that Data Loader uses to match existing records, such as AccountNumber, Id, Email, or SKU.

AI-assisted Data Loader helps you prepare Data Loader import and update jobs from source files. Describe what you want to load, upload your source file, review the suggested mappings and validation results, and submit the job after you confirm the changes.

From Zuora tenant, open **Data Loader** and select the **Zuora AI-assistant**.

From the **Chat** tab, describe the data operation that you want to perform.

For example:

* Import new accounts with billing contacts
* Update existing contacts by email

> **Note**
>
> The AI assistant uses your description to identify the object, operation, match key, and related-object handling.

Upload your source file.

The planned formats include CSV, Excel, JSON, and JSON Lines. The AI assistant normalizes the file and displays a tabular preview before mapping.

Review the target object, operation, match key, and related-object interpretation identified from your request and source file.

* Object: Account, Contact, Product, or another supported object.
* Operation: Import or create or update.
* Match key: Field that identifies an existing record during an update.
* Related Objects: Related records that the job handles together or in separate jobs.

Review and correct the field mappings.

The AI assistant maps source columns to tenant fields, including custom fields. You can ignore an unmapped column, map it to a custom field, or override the suggested mapping.

> **Note**
>
> For repeat loads, you can save the approved mapping as a reusable template.

Review and correct validation issues.

Data Loader validates the source data against the Zuora schema. Review and correct validation issues before you submit the job.

Validation can identify issues such as:

* Missing required fields
* Invalid picklist or enum values
* Incorrect date, currency or email formatting errors
* Duplicate update match keys
* Missing referenced records
* Custom-fields type mismatch

The validation panel can categorize issues by severity, such as blockers and warnings. Depending on the release, you might be able to export a list of corrections or ask AI to apply bulk transformations.

Confirm and submit the job.

Confirm that the selected operation matches your intent and submit the job. The submitted job appears in the standard Data Loader job list.

Select **Refresh** to update the job status. Monitor the job as it progresses from **Submitted** and then to **Completed** or **Failed**.