Automate business processes with the power of Machine Learning and AI?

We know Artificial Intelligence (AI) is more accurate than we humans are. For example at image recognizing. And though AI is not yet 100% accurate, they recognize images with smaller amount of errors than we humans do. What I’m trying to say, is that AI has a great potential for improving our lives and businesses.

Unfortunately, the reality is that currently more than half of the machine learning solutions developed fail to make it to production. We also know that half of the organisations struggle with the complexity of AI, machine learning, data science and so on. On top of that it is estimated that in the coming years there will be a huge shortage of data scientists. So if organisations want to do something with AI in the near future, finding a data scientist is going to be a problem. To help solving this issue SAP offers their AI Business Services.

What are the SAP AI Business Services?

Imagine there were prebuilt machine learning services we can use to solve a specific problem found in many different business processes, completely eliminating the need for a data scientist.

Sounds quite good, right? Well that’s exactly what the AI Business Services provide us with. They are reusable services that allow us to automate, optimize processes and enrich the customer experience across the intelligent suite. Currently SAP has six AI Business Services. I would like to go over each one of them in this blog, and give an example and the business benefits.


1. Automate tickets with SAP Service Ticket Intelligence

Meet Billy. Billy is a service agent for a big utilities company. Billy is first level support. He receives about a hundred emails a day and the phone also rings non-stop. His job is to identify the topics of the tickets he gets in, and then make the decision to forward it to second level support or try to solve it himself. Forwarding tickets to the right person takes up about 70% of his time. Leaving only little time for him to resolve tickets himself, or give a quick enough answer to his impatient customers.

Meet Billy…

A daily scenario looks something like this. He receives an email from a customer who says he has received an incorrect invoice. This is not something Billy is able to help with, so he has to forward the ticket to second level support. He doesn’t have enough time to give the customer a outstanding customer service experience because he is already being called by the next customer while the emails keep piling up in his inbox.

What if there was a way to make Billy’s job easier? The SAP Service Ticket Intelligence can automate the ticket classifying process. The service automatically identifies the ticket category, and then it will route the ticket to the correct person accordingly. Service Ticket intelligence can also recommend similar completed tickets to the employee who has to process the ticket. This can be useful because he can adapt the solution to his customer. Or if the problem is similar enough he can just copy the solution. Either way this ensures that the efficiency of processing tickets gets a huge boost. But this also drastically reduces the repetitive task of classifying tickets. Giving Billy much more time to focus on actually answering customer tickets instead of dispatching them and distributing them all day. The benefits are speeding up service response times, increasing problem resolution and ticket closure rates and increasing overall customer satisfaction.

2. Win time and reduce errors with SAP Document Information Extraction


This time, it is Billy’s job to go through hundreds of invoices in PDF format. And manually copy the data into the system, and in the corresponding fields in software of course. This takes him three days to complete and he hopes he doesn’t make any mistakes. This is a great use case for the SAP Document Information Extraction Service. The service is able to identify the information on the invoice and transfer it to the corresponding fields in the system automatically. In other words it’s able to digitalise unstructured business documents. The benefits are drastically reducing manual efforts, reducing costs, increasing document productivity and reducing errors.

3. Document Classification

Meet Emma..

Meet Emma. Emma works for a big insurance company, and like Billy it’s her job to take care of the enterprise mailbox. However, in the insurance industry they also have to deal with a large amount of paper documents. These documents are sent by post and they can be of many different types for example: Invoice, dunning letter, sales order, complaint or request. Its Emmas job to scan these documents into the computer and manually dispatch them to the right department to take care of.

Unlike Billy’s case we don’t need to extract all the relevant information or enrich the extracted information with additional information. The documents just have to be classified and dispatched to the correct department.

Well you might have guessed it, this is a great use case for the Document Classification Service. You can train the model to fit the customers exact needs and it completely automates the process for you. The benefits are minimising repetitive tasks, minimising errors, faster document processing which will lead to faster and more accurate responses to the customer.

4. Data Attribute Recommendation

Meet Dave..

Meet Dave. Dave works as a truck driver for a big furniture manufacturing company. One day he has to cross the border to deliver a big shipment internationally. He approaches the first border but he gets stopped at customs because he has the wrong commodity code printed on his customs document. Now Dave has to sit and wait until office fixes the document and sends it to him.

So, what went wrong here? Whenever an employee creates a new material in the system he also has to identify the right commodity code, because this gets printed on the customs document. This code is supported by rules, and in this case it’s eleven digits long because it has eleven levels of structure. So, something went wrong creating the commodity code, stranding poor Dave at customs and delaying the shipment, which results in high costs for the company.

This is a great use case for the Data Attribute Recommendation Service. You can train, for example, the service to learn the logic of the eleven digit structured commodity code, leaving zero room for errors or inconsistencies. The benefits are reducing manual efforts of course, increasing the validity of commodity codes and decreasing additional labor costs caused by human error.

5. Invoice Object Recommendation

Imagine you have an invoice incoming without any reference. No purchase order, nothing you can use to identify what it’s for. A lot of time will have to be invested to manually insert the required information such as the general ledger account. Not to mention it’s a highly repetitive task and cost intensive as well.

The Invoice Object Recommendation is designed to solve this business problem. The service uses machine learning to search for the correct general ledger account and automatically insert the required information. Then it will perform a check to see if the correct account has been assigned and charged. So, the benefits are processing more invoices efficiently and faster, increasing quality and compliance of G/L accounts and cost objects and of course educing repetitive manual work.

6. Business Entity Recognition

Last but not least let’s look how Business Entity Recognition can also help out Emma. Imagine she receives an email from a colleague asking if she can check the status of invoice number 879768 sent on the 25th of October. Apparently the supplier still hasn’t received payment and would like to know when it was paid.

The Business Entity Recognition Service would be able to identify named entities in the email (unstructured text) and then automatically transmit them to the system. The benefits are it allows you to use automation, handling a higher amount of text documents more efficiently and faster, increasing quality and compliance of your text documents handling processes and of course reducing manual repetitive work allowing for members of your organisation to focus on more relevant tasks in their field of expertise.

Wrapping Up

So, regardless of what industry your business is in. It is clear that there is a lot to be gained with AI and machine learning. We’ve gone through a few use cases now but, there are many more to be realized using the AI Business Services. Is your company ready to innovate, reduce operational costs, increase efficiency or take customer service to the next level? Are you interested in what the SAP AI Business Services can do for your organization? Want to learn more? Please feel free to contact us.

Schrijf je in voor onze nieuwsbrief

En blijf op de hoogte van de nieuwste evenementen