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

# Managing Data

> View and maintain data sources and their metadata in AskTable — basic info, connection settings, field configuration, and training sets — to keep queries accurate and data secure.

After a data source is created, you can maintain its metadata and configuration on the data source detail page. Good configuration helps the AI understand your business field semantics and produce more accurate analysis.

## Open the data source list

In the left main navigation, click **More**, then click **Data Sources** in the data workspace sidebar to open the data source list.

The list shows a grid of cards. Each card includes:

* **Engine icon + name + engine badge** — the data source name and its database/file type
* **Status indicator** — whether metadata is available and whether a sync is in progress
* **Description** — the business description of the data source
* **Table count** (database types also show the schema count) and **last updated time**

Hover over a card to reveal a **···** menu in the top right:

* **Quick Chat** — jump to the conversation page and ask questions against this data source
* **Delete Data** — delete the data source

Click a card to open its **detail page**.

## Detail page header

The top of the detail page is the data source overview: a breadcrumb (**All Data / current source**), the engine icon, name, engine badge, description, and a metadata row — the copyable data source **ID**, schema / table / field counts, availability status, and creation time.

Action buttons sit in the top right:

* **Connection** (database types only) — open the connection edit dialog
* **Edit** — modify basic info
* **···** — a menu with **Build Value Index** and **Delete Data Source**

## Edit basic info

Click **Edit** in the header to open the basic info dialog, where you can change:

* **Data name** — the display name of the data source
* **Description** — the business description, which helps the AI understand the data's context

## Edit connection (database types only)

For database sources, click **Connection** in the header to edit the connection details: host, port, username, password, and database. Leave the password blank to keep it unchanged. After saving, AskTable re-tests the connection and re-syncs metadata.

## Detail page tabs

The detail page organizes content into tabs:

* **Data Overview** — always shown; the core area for metadata and field configuration
* **File Management** — shown only for file types (Excel / CSV)
* **Training Set** — always shown

So database types have two tabs ("Data Overview, Training Set"), and file types have three ("Data Overview, File Management, Training Set").

## Data overview and field configuration

Data Overview is the default tab and the **most important configuration area**.

The toolbar shows the "Data Overview" title and a sync status badge on the left; on the right are a **Reselect Tables** button (database types only, to add or remove synced tables) and a **···** menu (export / import metadata).

### Page layout

* **Left tree navigation** — organized by schema → table; click a table name to view its fields on the right
* **Right side** — the selected table's **table description** at the top (a blue-bordered block; click the edit icon to modify), with the field table below

### Field configuration

The field table contains these columns:

| Column       | Description                                                                                                                                                                                                          |
| ------------ | -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| **Column**   | The original field name in the database                                                                                                                                                                              |
| **Note**     | The field's business description. An icon on the left marks its source: a database icon (original database comment), a robot icon (AI-generated), or a pen icon (human-annotated). Click the edit icon to change it. |
| **Hidden**   | A toggle. When hidden, the AI ignores this field — useful for passwords, internal IDs, and other sensitive or irrelevant fields.                                                                                     |
| **AI Index** | A toggle. When enabled, the AI builds a semantic index over this field's values to improve natural-language query accuracy. Click the info icon to view index statistics.                                            |
| **Masking**  | A dropdown. Mask fields containing private data — supports person name, phone, ID number, bank card, and email.                                                                                                      |
| **Updated**  | When this field's configuration was last modified                                                                                                                                                                    |

For details on each, see [Table and Field Notes](/en/concepts/data/annotations), [Field Hiding](/en/concepts/data/hidden-fields), and [AI Index](/en/concepts/data/ai-search).

## File management (file types only)

For Excel / CSV sources, the detail page has a **File Management** tab:

* **File list** — shows uploaded files; supports preview, download, and delete
* **Upload new files** — drag and drop or click to upload `.xlsx`, `.xls`, `.csv` files; multiple files supported
* After making changes, click **Save Changes** to apply

## Training set

The Training Set tab manages question-and-answer training data specific to this data source. Each entry contains:

* **Question** — a natural-language question
* **SQL** — the correct SQL statement
* **Source** — `Manual` or `Feedback` (collected from likes in conversations)
* **Created time**
* **Active** — a toggle controlling whether the entry takes effect

Training data helps the AI learn specific query patterns and business terms, improving answer accuracy.

## Import / export metadata

In the **···** menu on the data overview page:

* **Export Metadata** — export field descriptions, field visibility, and training data to a JSON file
* **Import Metadata** — import a JSON file to overwrite field descriptions and visibility, and append training data

## Delete a data source

Click the **···** menu in the header and choose **Delete Data Source**. After confirmation, the data source and all its configuration are permanently removed. Deleting a database source does not affect the actual data in the database.

## Next steps

* [Table and Field Notes](/en/concepts/data/annotations) — add business descriptions to improve AI accuracy
* [Field Hiding](/en/concepts/data/hidden-fields) — control which fields are visible to the AI
* [AI Index](/en/concepts/data/ai-search) — build a semantic index over field values to improve query accuracy