You can use the File Connector APIs to sync .txt, .md, .doc, .docx, .csv, .xls, .xlsx, .html, and .pdf files when using knowledge for Genesys AI-powered experiences.

The File Connector APIs support the following features:

• Full synchronization: Perform a full synchronization, in which all previously uploaded files are removed before the new file set is processed.
• Incremental synchronization with delete support: In addition to adding and updating files, incremental synchronizations support deleting previously uploaded files. This capability enables you to keep your knowledge source fully synchronized using a single synchronization operation.
• Abort synchronization: Abort an ongoing synchronization before indexing begins. This feature provides greater control over the synchronization process.
• Support for filenames with spaces: Upload and synchronize files that contain spaces in their filenames.
• Bulk upload support: Upload multiple files at once using a ZIP file, reducing the number of API calls required for large content updates.

These features simplify integration scenarios and improve synchronization efficiency for customers managing large content repositories.

A full synchronization process includes the following four phases:

1. Creating a knowledge source (once per source)
2. Starting a sync round (once per sync round)
3. Requesting upload URLs (once per file)
4. Marking the synchronization as completed (after all uploads)

Create a knowledge source

Create a source once, then reuse its sourceId for all future synchronizations.

Endpoint

POST /api/v2/knowledge/sources

When to call

Initial setup (or when provisioning a new source).

Response

Returns the created source details, including the sourceId.

Start a sync round

Before uploading any files for a sync round, you must create a new synchronization.

Endpoint

POST /api/v2/knowledge/sources/{sourceId}/synchronizations

Path parameters

sourceId — ID returned from the “Create a knowledge source” step.

When to call

At the start of every sync round.

Response

Returns synchronization details, including the synchronizationId.

Request upload URLs 

For every file you intend to upload as part of the synchronization, request an upload URL from the uploads endpoint.

Endpoint

POST /api/v2/knowledge/sources/{sourceId}/synchronizations/{synchronizationId}/uploads

Path parameters

• sourceId — knowledge source ID
• synchronizationId — ID returned from the “Start a sync round” step

When to call

Once per file you plan to upload.

Response

The API returns a URL to upload the file content (typically a pre-signed URL or similar mechanism, depending on implementation).

Expected flow per file

1. Call the uploads endpoint to obtain the upload URL.
2. Upload the file to the returned URL.
3. Repeat for all files in the sync round.

Mark the synchronization as completed

After all files have been uploaded successfully, mark the synchronization status as completed.

Endpoint

PATCH /api/v2/knowledge/sources/{sourceId}/synchronizations/{synchronizationId}

Path parameters

• sourceId — knowledge source ID
• synchronizationId — synchronization ID

When to call

Only after you have finished uploading all files for this synchronization round.

Behavior

Sets the synchronization status to completed (and signals the system that uploads are done).