Goodbye SAP Predictive Analytics - Welcome your be...
Technology Blogs by SAP
Learn how to extend and personalize SAP applications. Follow the SAP technology blog for insights into SAP BTP, ABAP, SAP Analytics Cloud, SAP HANA, and more.
When I first started working in customer advisory four years ago my main responsibility was to show customers our solution SAP Predictive Analytics. I explained them how to easily build predictive models without coding; how to discover hidden insights and relationships in your data, from which you can make predictions about future events. And I loved them; both the tool and my job. The good news is I still love my job but for the on premise solution SAP Predictive Analytics a lot has changed in the last 4 years. Due to new requirements from our users, it is deprecated and will run out of the mainstream maintenance phase end of 2024 (you can find the details on that and on the Priority-one support option here). This makes me a little sad and I wanted to say goodbye with this blog post to this great tool and also help others to carry on without this tool and experience the new and more modern alternatives SAP offers.
We live in fast changing times and our software solutions adapt quickly. The fact that predictive analytics applications are worthwhile, and their use brings measurable added value, doesn’t need to be discussed these days. Everyone talks about machine learning, predictive use cases, data science, data mining, advanced analytics-just to mention a few buzzwords. Many new job-roles developed in this field and to suit the individuals needs a total of three solutions will support you with your predictive scenarios:
First, more user groups and not just data scientists requesting access to predictive results. That is way we wanted the business user to get easy access to it. The business user is not satisfied with predictive result on their own. They want them to be easily consumable, improve their planning and to support them with their decision making. This was the main motivation to include some features from SAP Predictive Analytics into SAP Analytics Cloud. This features are called Predictive Planning and Smart Predict. The business user can build their own predictions as they are used to from SAP Predictive Analytics with a few clicks. Furthermore, SAP Analytics Cloud comes with a much nicer and a more simplified UI that is well fitted to the needs of the business users (let’s be honest SAP Predictive Analytics’ UI was old fashioned and ugly). SAP Analytics Cloud offers not only predictive features it also offers Business Intelligence and Planning.This is great since now you have everything in one tool and you can directly visualize your predictive results, integrate them in your dashboard and use them to improve your planning. A big improvement to the old world which only offered predictions.
The second option focuses more on our developer community. In SAP HANA we still provide the Automated Predictive Library, this library corresponds to SAP Predictive Analytics as well was the SAP Analytics Cloud Smart Predict engineand allows developers to quickly create and deploy predictive models in SAP HANA. Now, our developers can use not only SQL but also Python or R to call the Automated Predictive Library directly on SAP HANA. So, they can use their beloved coding environment to call the Automated Predictive Library Algorithms and let them run on the in-memory database SAP HANA.
Of course, the world does not only consist out of SAP even when would wish that. The third option helps you when the data is not stored in a single SAP HANA data base but is distributed across different system. With SAP Data Intelligence you can orchestrate and manage your data in such a “historically grown” IT landscape. Furthermore, you can deploy machine learning and predictive analytics use cases and, in this context, also call the Automated Predictive Library in SAP HANA or deploy your own R and Python code.
Any breakup offers new opportunities and I hope these new options help you with your separation pain from SAP Predictive Analytics. It helped me a lot to heal my broken heart. Just pick your best fit solution and celebrate the new opportunities and if you still feel sad you can of course contact me for more support.
I have been working as an academic in statistical tools like Minitab and Tableau. I also have some working knowledge of 'Microsoft Azure' where I can import data, perform ETL, create and train a model, followed by consume the trained model for future predictions. Within Azure I had the choice of various out of the box regression and classification algorithms, which I can pick and choose depending on the scenario.
Is there any equivalent in SAP ? I thought it would be predictive analytics but from your post it feels like SAP predictive analytics is phased out. Any guidance on where to start with building models and training them ?
Very well explained. I can understand the pain as i am also from the dinosaur era. joking....from the time when the user requirement's were more around results as that was the technology supported. Now with Tehnology advances and more augmented features built into Analytics , SAC, Data Intelligence supported by the HANA PAL/APL algorithms would really help business get the most of both worlds.
HI sarah.detzler - I am new to this topic of Predictive Analytics and while we deployed Fiori apps which use predictive analysis ( e.g F2012 ), I am not sure if I these apps would be ok to run or not. Does the maintenance policy of PAi have any impact on these Apps life in any way ?
This are great additions to the list. Just one remark from my side the PAL as a HANA native library will not be replaced by ISLM. You can still use PAL it independently on HANA. ISLM as our intelligent Scenario Lifecycle Management is a nice framework to use the PAL or APL and easily integrate integrate ML Scenarios for example in your S/4Hana System. Just to avoid misunderstandings.
All your blogs are very informative. Thank you for doing this.
Quick question - So, the SAP Predictive Analytics that is going away ( end of 2024 as Antonine Chabert mentioned as well), does not refer to the "Predictive Scenarios" in SAC - correct? I am assuming this will continue in SAC?
