Microsoft Cognitive Service - OCR

anyone has experienced the integration between the UiPath and iOCR from Microsoft Cognitive Services. Can share it with us?

we need the OCR to scan the unstructured input scanned documents.
What is the flow?
The robot installs the scanned document to a shared storage, then the Orchestrator calls the API of the OCR service.

the returned result of the service (json, xsl,…)
should be shared again by the robot?

Are there anything the Orchestrator to be integrated with SDK of Azure Cognitive Services?

according to Microsoft:
“No components must be installed on “clients” = robots machines. The use of the services is resolved with a configuration at the Orchestrator level that will incorporate the access library (SDK or direct access according to the cases) and that will guarantee the execution of the Azure services according to the needs of the use cases that you will implement”

Where can I find the documentation about this configuration in Orchestrator to integrate with the SDK or call directly the services?