Scope and Functionality of Machine Learning and Predictive Analytics with SAP S/4HANA
Part 4 of the blog series:
Continuing the blog series, I would like to now discuss the scope and functionality of the Machine Learning and Predictive Analytics leveraged with SAP S/4HANA. A quick re-cap of my earlier blog on the various approaches and process flows can be found here.
Ever since the inception of SAP’s flagship product S/4HANA in 2015, machine learning and predictive analytics technologies have been embedded to leverage additional benefits for the operational user and the business users. The picture below highlights the SAP strategy towards an intelligent enterprise.
Let us now dive into the machine learning and predictive analytics offerings in the S/4HANA digital core. As of this writing in early January 2020, there are around 40 use cases in the S/4HANA digital core that are embedded with predictive analytics algorithms or consuming machine learning services. We will look into some of these functionality in detail. The evolution of the intelligent core started by embedding predictive algorithms into some of these S/4HANA business processes across the different lines of businesses such as Finance, Procurement, Produce and Manufacturing, Sales and distribution, Portfolio and Project management etc., Ofcourse this is now culminating into enabling intelligence into the End-to-End processes across the different lines of businesses and industries which help the customers solve their problems in a more realistic way and see how SAP’s AI technologies are working in tandem. Here is a link to a great blog that explains the infusion of AI in the context of end-to-end processes, for eg., the order-to-cash process in the context of a sales manager.
Before I blog about the end-to-end processes and the context of how AI is infused, I would like to take you thru’ SAP’s journey into the Intelligent enterprise. Like I said before, there has been a lot of emphasis and efforts put into developing predictive intelligence into the LoBs such as finance and procurement with a chunk of the use cases embedding machine learning and predictive analytics. Ofcourse, a few more use cases in SD and manufacturing LoB have seen investments as well to enable predictive intelligence into the S/4HANA business processes too.
The beauty of SAP functionality is the way it is structured, and with the new S/4-centric approach, it is also easier to find and implement the required functionality. While the functionality is organized in terms of scope items with each function being assigned a 3 letter acronym, you could easily activate or de-activate picking up the right scope items. Most of the machine learning and predictive analytics functionality built with SAP S/4HANA is assigned a scope item.
One of the first scenarios relate to Cash application where the system would automatically match incoming payments with open receivables. Here the application takes advantage of machine learning to enable effortless set-up and to increase automatic matching rates. Furthermore with another machine learning service, unstructured payment advices / remittance advices could be automatically extracted to match customer payments and advices. Following are some of the use case functionalities that are available in the Finance line of business:
- Cash Application (1MV) – a quick look into the scope of this functionality
- Payment advice extractor
- Goods receipt / invoice receipt monitor status approval
- Payables line item matching
- Intelligent accrual recommendation
- Detection of abnormal liquidity items
- … and more
Each of the above machine learning services have been enhanced in the last 2 years to provide more functionality and details for the Finance accountants and Cash managers etc.
Procurement is another line of business where a huge investment has been made to enable predictive intelligence into the SAP S/4HANA digital core. One of the first use cases here focused on identifying contracts that are due for expiry and provide alerts so that better planning and contract re-negotiations could happen by saving huge costs for the procurement divisions. Further more, a lot of efforts were put in creating intelligence into the shopping cart so that free text items and material groups could be identified automatically based on past behaviors and added into the catalogs for saving costs. Some of the use cases that are already available in procurement line of business:
- Contract Consumption (1QR) – a sneak peek into the scope of this functionality
- Proposal of new catalog item
- Proposal of material group
- Proposal of options for materials without purchase contract
- Propose resolution for invoice payment block
- Intelligent approval workflow
- Image based buying
- Supplier delivery prediction
- … and more
I will focus on the topic of Procurement and the intelligent end-to-end scenarios in that context in a seperate blog. Most of the use cases in the procurement LoB focused on the Purchaser role – predominantly the Operational Purchaser and a few scenarios were geared towards the employee procuring items.
Soon to follow is the SD line of business where some realistic use cases around the probablity of a quotation conversion to sales order, sales delivery performance prediction and sales forecasting across different parameters have been developed and embedded into the S/4HANA Sales business processes. Following is a list of some of those use cases already delivered:
- Quotation Conversion probability rate (2YJ) – link to the functionality and scope
- Sales Forecasting
- Delviery performance / delivery in time
- Sales performance prediction
- … and more
We shall look into a detailed flow of some of these use cases in our upcoming blogs.
Finally, the Manufacturing Line of business came into it’s own with investments to help the manufacturing and production plants to benefit the AI wave by incorporating Machine Learning services and Predictive analytics functionality. One of the first use cases helped the ware house clerks and inventory managers to predict delay for stock in-transit between different locations and later extended this by identifying the slow and non-moving stock. This greatly helps the inventory managers to plan and replenish the stocks accordingly with huge bottom line savings for the companies.
Some of the released use cases in the manufacturing and production line of business include:
- Predict delay in stock in-transit (20N) – overview of the functionality
- Demand-driven replenishment
- Defect code proposal (incl. text recognition)
- Early detection of slow and non-moving stock
We will get into the specifics (in our upcoming blogs) of how some of the use case functionality has been embedded into the S/4HANA business processes.
Ofcourse, there is a lot more functionality developed across the S/4HANA domain to infuse predictive intelligence and make more sense for the end users in the enterprise space.
In my upcoming blogs, we shall discuss some of the use cases developed by leveraging the different approaches we talked about in the earlier writings!
Here are some quick links to the blogs in this series to give you a complete understanding of how Predictive Intelligence is infused into SAP S/4HANA.
- Part 1 – Leveraging Predictive Intelligence with SAP S/4HANA
- Part 2 – Architecture and deep-dive of the different approaches around Predictive Intelligence
- Part 3 – Process flow leveraging Machine Learning and Predictive Analytics
- Part 4 – Scope and functionality in the context of an end-to-end process leveraging ML (this blog)
- Part 5 – Activating machine learning functionality for SAP S/4HANA
- Part 6 – Building ML into the digital core of SAP S/4HANA (Embedded ML)
- Part 7 – Enhancing the digital core with ML Services (Side-by-Side ML)
- Part 8 – Extending the digital core by leveraging ML with SAP Analytics Cloud
- Part 9 – ML Extensions to SAP S/4HANA processes
- Part 10 – Leveraging Machine Learning with the ISLM Framework
- Introducing the book – Implementing Machine Learning with SAP S/4HANA
Happy predicting the future!!