Configuring Machine Learning Service

Work Instructions: Configuring Machine Learning Service for the Reconcile GR/IR Accounts App

This document provides detailed step-by-step work instructions for configuring the machine learning service for the Reconcile GR/IR Accounts app in SAP. Follow these instructions carefully to ensure a successful setup.

---

Prerequisites

Before starting the configuration process, ensure the following prerequisites are met:

1. GRIR Business Service Activation:

   • Ensure the GRIR business service is active in your SAP Business Technology Platform (SAP BTP) account.

2. SAP BTP Sub-account Setup:

   • Create a sub-account by following the instructions in Create a Subaccount.

   • Manage service quotas using the Cockpit as described in Configure Entitlements and Quotas for Subaccounts.

   • Create spaces in your sub-account by following Create Spaces.

3. Separation of Instances:

   • Create separate sub-accounts for testing/integration/quality/pre-production and production use.

   • Assign the testing instance to the testing sub-account and the production instance to the production sub-account.

4. Issue Resolution:

   • For setup issues, raise an incident in BCP under component CA-ML-FCA-IGR.

---

Configuration Steps

1. Create Service Instance and Keys

Procedure:

1. Open your SAP BTP space.

2. Navigate to Services > Service Marketplace.

3. Select the Data Attribute Recommendation (DAR) tile.

   Note: The GRIR machine learning service has been replaced by DAR. Refer to SAP Note 3238031 for more details.

4. Create a new service instance:

   • Choose New Instance.

   • Select the plan Intelligent Accounting.

   • Enter a desired instance name.

   • Choose Create.

5. Generate the service key:

   • Under Service Keys, select Create Service Key.

   • Record the generated oAuth credentials.

6. Note down the following communication settings from the service key JSON:

Setting	Location in JSON
Host Name	url
OAuth 2.0 Endpoint	uaa.url
OAuth 2.0 Token Endpoint	uaa.url + "/oauth/token"
Username	uaa.clientid
Password	uaa.clientsecret

---

2. Create RFC Destinations

Procedure:

1. Run transaction SM59 to create an HTTP connection to an external server:

   • Connection Type: G

   • Example Destination: DESTINATION_PROPOSALS

   • Path Prefix: /

   • URL: Use the URL from the service instance created in the previous step.

2. Run transaction OA2C_CONFIG:

   • Create the OAuth client profile FIS_GRIR_ML.

   • Enter the following details from the service key:

     • Client ID

     • Client Secret

     • URL

3. Configure access settings:

   • Client Authentication: Basic

   • Resource Access Authentication: Header Field

   • Grant Type: Client Credentials

4. Run transaction S4ML_GRIR_SETUP:

   • Specify the RFC destination: DESTINATION_PROPOSALS.

   • Activate the Save checkbox to save the configurations.

---

3. Schedule Training Job

Procedure:

1. Access the training job scheduler in SAP.

2. Define the job parameters and schedule the training job.

3. Verify the job execution and log any errors for troubleshooting.

---

4. Schedule Monitor Job

Procedure:

1. Access the monitor job scheduler in SAP.

2. Configure the job to monitor GRIR account reconciliations.

3. Schedule the monitoring job and ensure it runs periodically.

---

5. Schedule Inference Job

Procedure:

1. Access the inference job scheduler in SAP.

2. Define the parameters for inference processing.

3. Schedule the job and validate its functionality by reviewing the output.

---

Related Information

• Create Service Instance and Keys

• Create RFC Destinations

• Schedule Training Job

• Schedule Monitor Job

• Schedule Inference Job