Knowledge Store

Knowledge Store is a database-like feature for business reference data, such as vendor names, employee records, office addresses, or product codes. Hyperscience can use this data in Custom Supervision tasks, flows, and integrations.

Use Knowledge Store when people or automated processes need to choose from the same trusted data. This helps prevent typing errors, reduces manual work, and keeps processed data consistent.

Accessing this feature

Your access to the feature described in this article depends on your license package and pricing plan.

To learn which features are available to your organization and how to add more, contact your Hyperscience representative.

How Knowledge Store works

Knowledge Store organizes data into collections, items, and properties.

A typical workflow has four steps:

1. Prepare trusted business data in a CSV file or manage it through the API.

2. Store the data in a Knowledge Store collection.

3. Configure a Custom Supervision task, flow, or integration to retrieve the data.

4. Use the returned items to validate or enrich processed data. For example, a keyer can search for and select an employee instead of typing the name.

Before you begin

Permissions

The built-in Knowledge Worker Permission Group includes View Knowledge Store, but it does not include Edit Knowledge Store. For the permissions included in each built-in group, see Permission Groups.

Navigate the Storage page

To open Knowledge Store, click Storage in the left-hand sidebar. The Storage page lists the collections available in your instance.

Limits

Knowledge Store does not require every item in a collection to have the same properties. For predictable search results and table columns, use the same structure for all items in a collection.

View collections and items

• In the Storage section, click a collection to enter the Storage details page.

• The item-table columns are based on the most recently updated item in the collection. If items use different properties, the table might not show every property that exists in the collection.

Create or update a collection

Import a CSV file to create a collection or update its items. An import adds items whose external_id is new and updates items whose external_id already exists in that collection.

Prepare complete update rows

When an import updates an existing item, the values in that CSV row replace the item's stored properties. Include every property that the item should keep.

Prepare the CSV file

• Add one row for each item and one column for each property.

• Include an external_id column. Place it first so the file is also consistent with Knowledge Store exports.

• Use a unique external_id for each item that you want to update later. An external_id must be unique across Knowledge Store, not only within one collection.

• You can leave an external_id value empty. A later import cannot match that item for an update and creates another item instead.

• Do not use duplicate column names or duplicate non-empty external_id values.

For example:

Import the CSV file

1. Go to Storage, and click Import.

2. In Apply Collection Definition, select an existing collection or enter a name for a new collection.

3. Drag the CSV file into the dialog, or click Choose Files and select it.

4. Click Upload.

5. Review the confirmation or error message. A successful import shows the number of rows processed.

6. Click Done.

Export a collection (v43.2 and later)

Export an entire collection when you want to review or edit its items offline, re-import them, or move the collection to another environment. The export is a CSV file that is compatible with the Knowledge Store importer.

1. Go to Storage, and open the collection.

2. Click Export.

3. When the export is ready, open the Notifications menu.

4. Click Download.

The exported file starts with external_id and includes every property found in the collection. Empty cells appear when an item does not have a property.

Split large exports before re-importing them

An export can contain the entire collection, but each imported CSV file can contain no more than 10,000 rows. Split larger exports into files of no more than 10,000 rows before you import them.

Delete a collection

Deleting a collection also deletes all of its items.

1. Go to Storage.

2. In the collection's row, click the trash can icon.

3. In the confirmation dialog, click Delete.

You cannot delete an individual item in the application. To delete individual items, use the Knowledge Store API.

Manage Knowledge Store data through the API

Use the public v5 API to create, retrieve, update, and delete individual items. API requests require the API Access permission. For authentication and other requirements, see Hyperscience API Overview.

In v43.2 and later, you can also retrieve an alphabetical list of collections and the number of items in each collection. This helps an integration discover available collections before it retrieves or manages their items. For endpoints and response formats, see the Knowledge Store API documentation.

Review Knowledge Store audit records (v43.2 and later)

Go to Administration > Audit Log to review Knowledge Store activity.

Audit records cover item creation and deletion, collection imports, exports, and deletions. In v43.2 and later, they distinguish queued and completed exports, record rejected export attempts, identify imported source files, and provide clearer details about deleted collections and items. For more information, see Audit Log (v39.1 and later).

Use Knowledge Store data in Custom Supervision

Add a Search component to a Custom Supervision task when keyers need to find and select an item from a collection. The component can return the selected item's properties to the task.

For configuration details, see Search components in the Flows SDK documentation.