In recent months you may have heard a lot about the Intelligent Enterprise and, how we are enabling customers to effectively use their data assets to achieve their desired outcomes faster – and with less risk. At the core of the Intelligent Enterprise is Machine Learning, which enables computers to learn from large amounts of data without being explicitly programmed. The world’s most relevant enterprise data is part of SAP’s system and business network. Our aspiration to make all its enterprise solutions intelligent, and help customers to leverage their data.
One of the key enablers of this digital revolution is the newest edition to the Predictive Analytics suite namely Predictive Analytics integator (PAi). PAi enables SAP Applications such as SAP S/4HANA to create and ship predictive use cases specific to their business directly to their customers. Customers can than easily manage the lifecycles of these models by retraining and debriefing these models directly within the applications.
Furthermore customer and partners can adapt predictive use cases, or create new ones based on their own business needs in SAP Predictive Analytics, and publish to the application with one click.
Fig 1. SAP predictive Analytics integrator
From S/4HANA Cloud Edition 1705 and S/4HANA On premise 1709 or above, we have been delivering new predictive content across many business areas out of the box using PAi. Our Domain Experts across Finance, Procurement, Sales, to name but a few, have been working closely with our Data Scientists to identify key business areas where predictive and machine learning techniques can improve efficiency, reduce costs, and automate time consuming business processes. Below we have listed some of the predictive scenarios already available for customers to use.
Fig 2. Predictive Scenarios live with SAP S/4HANA since 1705
If you want to know how you can get started with all this new predictive content you have come to the right place. In this Blog Series the PAi team and I will introduce the Predictive Scenario, and take you through how to get started, how to use what SAP is shipping out of the box, and how you can customize these Predictive Scenarios to meet your specific needs.
Blog 1: Installation and Configuration by Shekhar Chhabra
If you are an S/4HANA Cloud customer (1705+), there is no installation or configuration required. This is all taken care for you by SAP, so you can move straight to Blog 2.
However, if you are an S/4HANA On Premise customer (1709+), there are some steps required to get you up in running. In our first Blog Shekhar will take you through how to install the required components in SAP how to configure the SAP Netweaver elements.
Blog 2: Training and Activating the Models shipped by S/4HANA by Orla Cullen
Once PAi is successfully installed, I will take you through how you can start to train the models that are shipped out of the box with S/4HANA and how you can periodically review and retrain them directly within S/4HANA with our Fiori tiles.
Publish to PAi
Fig 2. Publish new Predictive Scenario from SAP Predictive Factory to SAP S/4HANA via PAi
Once you have been successfully using the Predictive Scenarios S/4HANA are shipping out of the box with PAi, you may already have some ideas of your own. In the next few blogs by Shekhar, Paul, and I, are going to start looking at creating your own predictive models in SAP Predictive Analytics, and publishing them back in to PAi. Shekhar will cover how to get started, and install Predictive Factory to be used with PAi. Paul will guide you through an approach to modeling for those new to the topic, and finally I will take you through a full worked example.
Blog 3: Installing and Configurating Predictive Factory and PAi by Shekhar Chhabra
Now that you are ready to start modellingwill need to install SAP Predictive Analytics, and connect it to the PAi application within your S/4HANA system.
Blog 4: Approach to Modeling by Paul O’Hara
So, now you have been using the existing Predictive Scenarios shipped out of the box with S/4HANA, but you have some new ideas for predictive use cases that you want to try out. In this Blog, Paul will take you through how to approach predictive modeling, and some of the things to keep in mind when building a new use case.
Blog 5: Defining the Target Variable by Paul O’Hara
Paul will walk you through one of the most important stages of a predictive use case, namely defining the target value to be predicted. It is critical that the target value is correctly defined, as it can have a significant impact on the output for a use case.
Blog 6: Worked Example Invoice Payment Forecast by Orla Cullen
In this blog we use the example of an Invoice Payment Forecast Predictive Scenario to pull together all the elements covered the previous blogs.
We will endeavour to release weekly episodes to this series, at every stage your feedback and suggestions for new topics would be much appreciated.