SAP S/4HANA Cloud Intelligent scenarios in Finance – #1 Machine Learning
The key characteristics for any Intelligent Enterprises would be increased automation, reduction of repetitive, non- value generating tasks and increased involvement in strategic decisions with more insights to action.
SAP is helping organisations to implement Intelligent ERP with the help of technologies like Machine Learning, Intelligent Robotic Process Automation and Situation Handling.
I would like to give you a brief about the key Intelligent Scenarios in SAP S/4HANA Cloud Finance across these technologies in a series.
- Machine Learning scenarios
- Intelligent Robotic Process Automation Scenarios
- Situation Handling scenarios
Let’s start with the Machine Learning in this blog.
Machine Learning is nothing but a set of algorithms which improve automatically through experience. It could work on the historical sample data and helps in making predictions or decisions. The SAP Leonardo ML foundation aims at providing an enterprise-grade platform for Machine Learning in the cloud.
I would like to introduce you to the first Machine Learning Scenario in Finance, i.e. SAP Cash Application.
The number of customer base or banks increases as and when the business grows and this increases the need for the account receivable team to be able to handle the high volume of payments. Though there are rule based solutions available, there would be a percentage of payments and invoices which require some manual interventions. This includes uploading of a payment advice received in an email, manual clearing of invoices and payments which is not cleared due to lack of details provided etc.
That is where SAP Cash application plays a role.
What is SAP Cash Application?
SAP Cash Application is a cloud solution that integrates with S/4HANA wherever it is deployed, cloud or on-prem. It is comprised of several intelligent services that facilitate automation in the cash application process including: automating the mapping of incoming bank statement items to open receivables items or accounts, automatic information extraction from unstructured remittance advice documents, and configurable auto clearing of confident proposals.
Fig1 SAP Cash Application
There are different services in SAP Cash Application like:-
- Receivable Line Item Matching
- Payment Advice Extraction
- Payable Line Item Matching
- Lockbox Processing
How does it work?
The model is trained with the historical clearing information which is sent from SAP S/4HANA to the SAP Cash Application cloud service and matching criteria can be derived. Training can be scheduled regularly to ensure changing behaviour is captured so the model can adapt.
Once after the model is trained, when new bank statements are received (most times daily), they will be sent to the service along with the open receivables, so the machine learning model can infer matching proposals. Proposals are returned to SAP S/4HANA and those that meet the configurable confidence threshold are automatically cleared for full automation. When there are multiple proposals for a payment they are presented to the AR accountant within the standard Fiori app they use today.
Payment advice documents received from the payer can be uploaded to the Manage Payment Advice Fiori app for automatic processing. The payment advice extraction feature within SAP Cash Application will accept PDFs including, unstructured scans, and use computer vision technology to read and extract information into the ERP. Once these payment advice documents are converted into structured information they are used to enhance payment clearing.
With such machine learning services, you can see an immense increase in the efficiency and reduction of errors in the finance department. The incoming payment processing will be much faster. Fig 2 below depicts the Account Receivable process with machine learning service automation.
Fig 2 AR process with automation
Now let’s look at the second scenario – Machine Learning for Monitoring of Goods and Invoice Receipt.
The GR/IR reconciliation process is an exception handling process for all purchase order items, where invoice receipts and goods receipts do not balance. As a G/L accountant and other involved parties, you need to understand the business situation at the level of a purchase order item and decide how to solve the situation.
Many companies which deal with large amount of purchase orders are familiar with this exception handling process. As part of the monthly, quarterly or yearly closing activities accountants have to check the GRIR accounts and have to make sure these balance out on the purchase order line item level. There are many reasons causing discrepancies between the goods receipt and the invoice receipt. For instance, we could have goods received with quantities deviating from the corresponding invoice receipt.
To resolve these kind of exceptions the accountant needs to collaborate with the different stakeholders such as accounting, logistics and the supplier. This used to be a very time-consuming activity. The accountant had to manually gather the information from different transactions and then manually contact the different stakeholders.
The ML service – Machine Learning for Monitoring of Goods and Invoice Receipt can be used in such circumstances. The ML service learns from the decisions made in the past and applies the learned knowledge to the new business situation, and proposes the next meaningful steps, the priority and root cause for each item.
Fig3: GRIR process
The Goods and Invoice Receipts in a normal case, run into the rule based clearing and if they match, the process continues and everything is fine. But if it doesn’t, then the Intelligent recommendation can be leveraged to provide a recommendation on how to proceed-Fig3. The employee can efficiently clear the accounts with these suggestions.
These intelligent recommendations simplify and accelerate the reconciliation process. The machine learning helps to automate accountants decisions with fewer human interactions and reduce the effort for non-value activities. Overall you can achieve a faster period end closing with higher accuracy.
Hope it helps you to understand the key machine learning scenarios in Finance.
You can find more informations on these scenarios from the below links as well.
In my next blog, I will brief you about the Intelligent Robotic Process Automation Scenarios in Finance.