2. Data Extraction with Cognitive Services

Azure Logic Apps used the Form Recognizer API from Azure Cognitive Services to extract key information (e.g., invoice number, supplier name, total amount) from the uploaded files.

Implementation:

• The blob URL was passed to the Form Recognizer API.
• Extracted data was saved in Azure SQL Database for validation.

JSON Output:

{

    "InvoiceNumber": "INV12345",

    "SupplierName": "ABC Supplies",

    "TotalAmount": 2500.00,

    "Date": "2024-11-30"

}

3. Validation Against ERP Data

The extracted data was validated against purchase order records in the client’s SAP ERP system using an HTTP connector. Any discrepancies were flagged, and the workflow sent an email to the finance team for manual review.

Logic Apps Configuration:

• HTTP GET request to SAP API for purchase order data.
• Compare the extracted data with the API response.

4. Approval Notifications

For invoices that passed validation, Logic Apps sent approval emails to designated approvers using the Outlook 365 connector. The emails included actionable links (Approve/Reject) to update the invoice status directly.

Example Email Content:

Subject: Invoice Approval Needed - INV12345

Dear Approver,

An invoice from ABC Supplies for $2500.00 is awaiting your approval. 

Please take action using the links below:

- [Approve Invoice](#)

- [Reject Invoice](#)

The response was captured via HTTP POST endpoints configured in the Logic App.

5. Payment and Final Notifications

Once approved, the invoice details were sent to the finance system for payment processing via the SAP ERP connector. Notifications were sent to all relevant stakeholders, including suppliers, using Microsoft Teams and email.

Business Benefits

1. Time Savings

The automated workflow reduced invoice processing time by 60%, ensuring faster payments to suppliers.

2. Enhanced Accuracy

By automating data extraction and validation, the Logic App minimized human errors in the approval process.

3. Scalability

The solution scaled effortlessly to handle peak workloads during high-volume periods.

4. Real-Time Visibility

Finance teams gained real-time insights into the status of invoices through Power BI dashboards, which were updated with data from the Azure SQL Database.

5. Cost Efficiency

By replacing manual processes with a serverless solution, the client reduced operational costs and improved efficiency.

Sample Logic Apps Workflow Diagram

Below is an overview of the implemented workflow:

1. Trigger: New invoice uploaded to Blob Storage.
2. Action: Extract data using Form Recognizer.
3. Action: Validate data with SAP ERP.
4. Condition:
   • If valid → Send approval email.
   • If invalid → Notify finance team.
5. Approval: Approve/Reject via actionable links.
6. Action: Update invoice status in ERP.
7. Notify: Send notifications to stakeholders.

Code Example: Logic App Designer JSON

Here’s an example of the Logic Apps Designer configuration in JSON:

{

    "definition": {

        "$schema": "https://schema.management.azure.com/providers/Microsoft.Logic/schemas/2016-06-01/workflowDefinition.json#",

        "actions": {

            "ExtractInvoiceData": {

                "type": "Http",

                "inputs": {

                    "method": "POST",

                    "uri": "https://formrecognizerapi.azurewebsites.net/extract",

                    "body": "@{triggerBody()}"

                }

            },

            "ValidateData": {

                "type": "Http",

                "inputs": {

                    "method": "GET",

                    "uri": "https://sapapi.company.com/validate",

                    "parameters": {

                        "InvoiceNumber": "@body('ExtractInvoiceData').InvoiceNumber"

                    }

                }

            },

            "SendApprovalEmail": {

                "type": "Office365.SendEmail",

                "inputs": {

                    "To": "approver@company.com",

                    "Subject": "Invoice Approval Required",

                    "Body": "<html>...</html>"

                }

            }

        }

    }

}