Categorizing Machine Learning Scenarios in SAP S/4HANA Utilities

Objective

After completing this lesson, you will be able to utilize Intelligent Scenario Lifecycle Management (ISLM) to optimize machine learning models and enhance operational efficiency in SAP S/4HANA Utilities.

Intelligent Scenario Lifecycle Management (ISLM) SAP S/4HANA Utilities

Intelligent Scenario Lifecycle Management (ISLM) is a framework that enables you to perform lifecycle management operations for machine learning scenarios. You can use ISLM as a self-service tool that allows you to handle various requirements of machine learning scenarios. You can train an intelligent scenario and use the trained model to get an accurate inference result.

Depending on the machine learning scenario associated with a business application, the scenarios can be categorized as follows:

Embedded: In this approach, a business application, for example, B. SAP S/4HANA runs in the same stack as the machine learning provider in SAP HANA with the analytics libraries SAP HANA Automated Predictive Library (APL) or SAP HANA Predictive Analysis Library (PAL). APL provides the data mining capabilities of an Automated Analytics engine for developing predictive modeling processes for business analysts. PAL offers advanced analytics algorithms for data scientists. These can be used to solve use cases of forecasts, trends, and so on.

Side-by-Side: In this approach, a business application, for example, B. SAP S/4HANA, running in a different stack than the machine learning infrastructure provider. Examples are SAP Data Intelligence, SAP AI Core, and SAP Business Technology Platform (SAP BTP) based services such as Intelligent Intercompany Reconciliation (ICR), Document Information Extraction, Data Attribute Recommendation, and Business Entity Recognition. Remote machine learning services can be used for sophisticated use cases such as image recognition, sentiment analysis, and deep learning for natural language processing based on neural networks.

Flowchart illustrating the development and consumption of intelligent scenarios in SAP S/4HANA. The development side includes roles such as ABAP Developer, who develops business logic, and Intelligent Scenario Owner, who registers and publishes scenarios. The consumption side involves roles like Business User, Business Administrator, and Technical Administrator, who utilize AI inference results, perform AI operations, and handle technical configurations respectively. Central components include SAP S/4HANA, Intelligent Scenario Lifecycle Management, and SAP HANA (APL, PAL). AI runtimes and services like SAP Data Intelligence, AI Core, and AI Business Service connect via AI API for operations and dataset management. The Data Science Developer develops machine learning scenarios using AI Core and SAP Data Intelligence.

Use of ISLM for forecasted days to payment in SAP S/4HANA Utilities.

Diagram of SAP S/4HANA intelligent scenarios showing inputs, process, and outputs. Inputs include Business Partner and Contract Account Data, Event History, and External Events. The process depicted involves an Intelligent Scenario with components such as Target Variable, Feature Logic, ML Model Setup, and ML Model Operations (Training/Prediction). Outputs generated are Score (number of days), Influencing Factors, and Automated Clustering.

Intelligent collections based on customer payment behavior (FI-CA, SAP behavior insights)

  • Based on ISLM

  • Less cost of collection management

  • Higher cash flow

With this predefined predictive scenario, you as a data scientist or machine learning expert can enable your users to predict the risk of late payment for invoices. After activating the scenario, the Risk of Late Payment field supports collection specialists in their daily work. The prediction of the risk of late payments is based on historical data and uses intelligent scenario lifecycle management.

Solution information:

  • Predictive Scenario FIN_COL_RSK_LTE_PMNT

  • View the risk of late payments for your invoices app

  • Integration with Collection Management (Dunning by Collection Strategy)

  • The prerequisite for use is that the collection strategy is used.

  • Manage Invoice Prioritization Rules App

SAP Help Portal: Predictive Scenario for Risk of Late Payment |

Implementing Chatbots and Virtual Consultants in SAP S/4HANA Utilities

Automatically Resolve Implausible Meter Readings Using AI

Challenge

  • As consequence of the energy transition, changing consumption patterns and more complex energy products increase the frequency and complexity of interactions between Utilities and end customers
  • High operational cost for customer service
  • Availability of customer service experts / resource scarcity
  • Risk of customer dissatisfaction and churn
  • Higher demands from customers towards customers service experience
  • Conventional chat bots are not aware of the individual customer data known to the Utility

Solution

  • A conversation-driven self-service agent for end-customers that represents an alternative to today’s self-service portals by communicating in natural language and providing responses in the most appropriate media (text, image, PDF, video).
  • The agent has knowledge about the individual customer’s meter-to-cash data in SAP S/4HANA Utilities and guides through the process tailored to their current situation, for example,
    • Proactive bill-shock prevention and consumption transparency
    • tariff recommendations, best product
    • update budget billing amount or installment plan
    • banking account details

Benefits

  • Become trusted advisor for your customer in the energy transition
  • Increase customer satisfaction and stickiness
  • Increase number of touchpoints with customers
  • Leverage upselling potential •Increase customer loyalty
The image is a process flow for addressing unusual consumption in SAP S/4HANA Utilities. Steps: System detects unusual consumption leading to high cost for the customer. System contacts customer through e-mail. Self-Service Agent clarifies with customer whether consumption is correct. Customer confirms higher consumption through new wall box. Agent proposes new tariff, which the customer accepts. Adjusted tariff is processed in SAP S/4HANA Utilities. Outcome: Customer is satisfied.

Business Value and Impact Overview Across Industries

Business ValueBusiness Impact ReferenceApplicable Industries

Reduce operational cost by offloading consulting work to AI based self-services

Increase customer satisfaction and loyalty

Utility company with 1 million customers, 10 percent of bills are suspect, half of which customers call the call center. Call duration: 25 minutes. Cost per hour: EUR 50:

1 million EUR savings per year

Utilities

(regulated, integrated and retailers)

Note

Check out these extra links if you want to learn more or deepen your understanding. They're packed with further information to enrich your knowledge.

Log in to track your progress & complete quizzes

Learning

SAP FacebookSAP YoutubeSAP LinkedIn