The Intelligent Scenario Lifecycle Management (ISLM) is new machine learning framework in SAP S/4HANA, available as of S/4HANA 2020. ISLM is the successor of the Predictive Analytics Integrator(PAi).
Table of Contents
What is Intelligent Scenario Lifecycle Management (ISLM)?
The Intelligent Scenario Lifecycle Management is new self-service machine learning framework in SAP S/4HANA, available as of S/4HANA On-Premise 2020 FPS00 and S/4HANA Cloud 2011. It enriches business process by leveraging machine learning and S/4HANA embedded analytics capabilities.
ISLM manages training, deployment, and activation for all intelligent scenarios using a consistent approach. An intelligent scenario is an ABAP representation of machine learning business specific use case.
It provides support for standard and custom developed ML scenarios including:
- Automated predictive library (APL)
- Predictive analysis library (PAL)
- HANA Embedded Machine Language Interface (HEMI)
- Remote scenarios based on SAP Data Intelligence
Intelligent Scenarios
There are 2 different approaches to enable machine learning (ML) to enrich SAP S/4HANA business processes. The following can be used to group machine learning scenarios that are tied to SAP S/4HANA business applications:
1. Embedded Machine Learning
In this approach, business applications and SAP HANA machine learning provider runs in the same stack which is SAP S/4HANA. Best suited for intelligent scenarios such as forecasting and trend analysis which require less demand for data, RAM, or CPU. This approach embeds Predictive Intelligence into the SAP S/4HANA business processes by leveraging the analytics libraries provided by SAP HANA.
- APL (Automated Predictive Library) – APL exposes the data mining features of an Automated Analytics engine to make predictive modeling processes.
- PAL (Predictive Analytics Library)** – PAL offers advanced analytics algorithms
- HANA Embedded Machine Learning Language Interface (HEMI), a PAL library
2. Side by Side Machine Learning
Using this approach, SAP S/4HANA, a business application, and SAP Data Intelligence, a machine learning provider, operate on distinct stacks. The most advanced applications of machine learning can be performed remotely, including image identification, sentiment analysis, and deep learning for neural network-based natural language processing.
The machine learning models consumed by the SAP S/4HANA processes are built on SAP Business Technology Platform using
- SAP Data Intelligence
- SAP AI services
ISLM supports creation of new customer developed intelligent scenarios and its life cycle management using two SAP Fiori apps, Intelligent Scenario and Intelligent Scenario Management.
Intelligent Scenario Fiori App
The Intelligent Scenario app can be used to both display standard scenarios provided by SAP and create your own custom intelligent scenarios. The custom-built intelligent scenarios are developed in a mode that may be reviewed, revised, and published. It is the successor of the Predictive Scenarios app.
Intelligent Scenario Management Fiori App
The Intelligent Scenario Management Fiori app can be used for managing machine learning tasks of intelligent scenario like training, deployment, and activation. It is the successor of the Predictive Models app.
Initial Setup
To enable ISLM in SAP S/4HANA, the following initial setup and configurations need to be performed.
- SAP HANA Automated Predictive Library
- Application Function Library (AFL), which includes SAP HANA Predictive Analysis Library
- Predictive Analytics integrator (UMML4HANA 1) Delivery Unit
- SAP HANA database user or ABAP technical user roles with sap.hana.pai::ExecutePAI role
- Configuration
- Set up the OData Services
- Set up the bgRFC Queue
- Configure Business Roles for the Backend
- Configure Business Roles for the Frontend
Summary
It is the framework for creating, managing, and executing custom developed machine-learning (ML) scenarios.