Installation

General

This document guides on installing and configuring the Form Recognizer Component.

Prerequisites

System and version

System	Version
Novacura Flow Server	6.14
Novacura Flow Studio	6.14
Connectors	Database, REST Service, Send E-Mail, Convert HTML To PDF
Azure Resources	Azure Form Recognizer Cognitive Service, Azure Blob Storage, Azure Table Storage

Azure Resources

Service	Link
Form Recognizer API Documentation	https://westus.dev.cognitive.microsoft.com/docs/services/form-recognizer-api-2022-08-31/operations/AnalyzeDocument
Microsoft Form Recognizer Studio Documentation	https://learn.microsoft.com/en-us/azure/applied-ai-services/form-recognizer/studio-overview?view=form-recog-3.0.0
Azure Form Recognizer Cognitive Service	https://azure.microsoft.com/en-us/services/form-recognizer/
Azure Blob Storage Setup	https://docs.microsoft.com/en-us/azure/storage/common/storage-account-create?tabs=azure-portal
Azure Blob Storage SAS Key	https://docs.microsoft.com/en-us/azure/storage/common/storage-sas-overview

Flow Connectors

IFS Connector

If you do not have a connector set up, you can see an example here.

HTML To PDF Connector

If you don't have a connector set up, you can see an example here.

Send E-Mail Connector

If you don't have a connector set up, you can see an example here.

REST Connector

REST - Azure Blob Storage

A REST Connector Project must be set up to utilize Azure Blob Storage.

Create a new REST Service Connector named "Azure Blob Storage v3" and import the configuration file "Azure Blob Storage v3.ncrcp." For more information see Import a REST Configuration.

There are some additional steps to do:

• Replace all the API key values in the connector

• Select “Save” to save the API Key

• Select “Save” to save the REST Connector Configuration

API values are coming from Azure Blob Service SAS URL

Review Microsoft Blob Storage's Documentation for additional information on how to utilize this service.

REST - Azure Table Storage

A REST Connector Project must be set up to utilize Azure Table Storage.

Create a new REST Service Connector named "Azure Table Storage v3" and import the configuration file "Azure Table Storage v3.ncrcp." For more information see Import a REST Configuration.

There are some additional steps to do:

• Replace all the API key values in the connector

• Select “Save” to save the API Key

• Select “Save” to save the REST Connector Configuration

API values are coming from Azure Table Service SAS URL

Review Microsoft Table Storage's Documentation for additional information on how to utilize this service.

REST - Azure Form Recognizer Custom Model API

A REST Connector Project must be set up to utilize the Microsoft Form Recognizer Cognitive Service.

Create a new REST Service Connector named "Azure Form Recognizer Custom Model API v3" and import the configuration file "Azure Form Recognizer Custom Model API v3.ncrcp." For more information see Import a REST Configuration.

• Enter a Base Address (example: https://westus2.api.cognitive.microsoft.com )

Replace this with the Endpoint from your previously created Form Recognizer Cognitive Service.

Review Azure Form Recognizer's Documentation for additional information on how to utilize this service.

Flow Properties

Before installing the workflows, it is required to create several Global Flow Properties within Novacura Flow Studio.

Flow Text Properties

Flow Property	Example Value	Description
AppOwner	ifsapp	IFS environment application owner (Schema).
OCR30_ToolUrl	https://formrecognizer.appliedai.azure.com/studio/custommodel/projects	Form Recognizer Studio website - Custom Model/Projects
OCR30_ModelTableName	FormRecognizerModels	The name of the table in the Storage Account to store the project-model key-value pairs
OCR30_ContainerName	formrecognizer-container	The name of the container that contains the uploaded BLOB files
OCR30_Environment	TEST	The table storage environment where this custom-model is used

Installation

Make sure that you have created all services, connectors, and Flow properties before importing the workflows.

1. Import Workflows

Import workflows in the file "Form Recognizer.wap".

For additional details see Import Workflows.

2. Setting Menu Roles

Together with the imported workflows there will also be a menu. Connect the menu to your roles to make them available for users.

For details see Setting Menu Roles

3. Publishing Workflows

Within the Server Contents Window right click on the Form Recognizer folder and select “Publish”. This process can take a few minutes but once complete, a Publish Successful message will be displayed.

If you receive any errors during the publication of the workflow. Please review the error detail and consult your support contact with the relevant information.

The workflows will now be available to execute on your chosen Flow client. For instructions on individual Form Recognizer functions, please consult the function individual documentation provided separately.

For details see Publishing Workflows