> ## Documentation Index
> Fetch the complete documentation index at: https://docs.k16solutions.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Import CSV data

> Upload CSV data, review the schema and sample rows, and assign access before publishing.

A DataX import turns uploaded CSV data into a catalog entity. Before you begin, you need **Manage data imports** permission, Lakehouse availability, a unique entity name, and an approved CSV file.

## Create the import

1. Open **Lakehouse** > **Data Imports**, then select **New Import**.
2. Enter **Name**, **Tags**, and **Description**. The name must start with a letter or underscore and use letters, numbers, underscores, or dollar signs.
3. Add one CSV under **Upload CSV file**. The creation form accepts one file; you can add more in the review page before publication.
4. Review **CSV parsing**, including **Field delimiter**, **Field enclosure**, **Encoding**, **Null values**, **Trim spaces**, **Skip blank lines**, and **Empty field as null**.
5. Select **Create and scan**. Keep the page open until the upload finishes and the import detail opens. Navigating away during upload cancels it.

## Review files, columns, rows, and access

The **Scan review** area takes you through **Files**, **Columns**, **Rows**, and **Data access**. Resolve errors before publishing.

| Step            | What to verify                                                                                                                                                                                         |
| --------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ |
| **Files**       | Confirm the uploaded source files and parsing settings. Add files with **Upload and scan**, or change parsing and select **Save CSV config** to scan again.                                            |
| **Columns**     | Review inferred types, edit column names, and apply the available constraints. Data types come from the scan and cannot be edited here. Select **Save columns** after changes and wait for rescanning. |
| **Rows**        | Wait for sample rows to load, then verify identifiers, dates, nulls, and column alignment. Sampling is required before publication.                                                                    |
| **Data access** | Review which roles can query the published entity. If you change selections, select **Save data access** and wait for it to finish.                                                                    |

A successful scan alone is not enough to publish. K16 also requires valid inferred types, loaded sample rows, and a review of **Data access**. Hover over a disabled **Publish** control to see the remaining requirement.

If a type is wrong, correct the source CSV or parsing configuration and scan again. Do not use column renaming to hide a parsing problem.

## Publish and verify

1. Confirm that all changes are saved and the review steps are complete.
2. Select **Publish**.
3. Wait for **Import published.** and the Catalog destination.
4. Review the entity's columns and query access, then run a small query to confirm the imported data.

After publication, the import review is read-only. Manage subsequent role access in **Settings** > **Data access**. Create a new import for replacement data rather than assuming a published import accepts additional files.

Deleting an unpublished configuration can leave uploaded files in workspace storage. Review the deletion message and your retention requirements before confirming.

For shifted columns, check delimiter and enclosure. For unexpected nulls, check null-value settings. For upload or scan failures, keep the exact error and import name for [Support](/support).