What’s in Store for 2020 with Machine Learning in Support

As we enter a new decade, there is much to reflect on in the area of emerging technologies.

It has not even been a full decade since technologies such as artificial intelligence (AI) and machine learning were introduced into the B2B support world, so we know we have only scratched the surface in terms of how these technologies can help improve business processes.

We sat down with Jens Trotzky, head of Artificial Intelligence Technology for Customer Success Services at SAP, to discuss what he and his team achieved over the last year in the development of automated support — and what they are looking forward to in 2020.

Q: What successes have you seen with AI and machine learning at SAP over the course of 2019?

A: You could say that 2019 was the first chapter of our ever-evolving journey toward an automated service and support function – one that is fueled by the Intelligent Enterprise, for the Intelligent Enterprise, and one that integrates AI and machine learning applications in support.

AI did its first real-world heavy lifting for SAP customers in 2019, allowing our team to observe various use cases within automated support and how they’ve impacted our customers in the early stages. Over the course of last year, we conducted a field test, put the intelligent machine out there, and gathered a lot of feedback from our customers, internal experts, and support engineers, which allowed us to continue to improve and build upon our intelligent and automated services.

Most prominently, we have seen the greatest success through the official go-live of incident solution matching via SAP ONE Support Launchpad. The service automatically connects users with the appropriate resources through natural language processing as they type, driving them to relevant answers to their technical questions faster. Through the collection of this interaction data, and fueled by AI and machine learning, the models are automatically trained and continue to improve. Customers who have used this service are extremely satisfied with the results of receiving the correct solution to their incidents so early in the process.

What has surprised you about the application of AI and machine learning in SAP’s support?

I am actually very happy about the stability of the cloud architecture we developed to deploy our AI and machine learning capabilities. Our AI applications are running on cloud-native applications that are able to scale exponentially, as needed, so we don’t need to invest any additional resources in order to cater to a growing customer base.

As we begin to see the implementation of our services among our customer base, we are learning that data is incredibly important in AI- and machine learning-driven applications. Not only have we learned that data quantity is important, but that data quality truly determines the business value we can provide our customers. While we’ve been able to gather vast amounts of data over the past year, we are working on getting access to additional streams of data to better streamline the interaction between the user interface and machine learning.

Where will you be focusing your efforts in 2020?

While 2019 marked the year we introduced semi-automated AI solutions to the services and support world, 2020 will revolve around testing the grounds for full automation, where AI will take on a completely independent business process step.

We will focus our efforts on perfecting our current solutions and rolling out more automated, personalized, and interactive features that provide our customers with the tailored solutions they need. Not only will we be introducing new services, branching out into the areas of predictive support, but we will also be bundling together previous standalone machine learning services to form an intelligent network for our customers in services and support.

Along with the deployment of these services to our customers, we will continue to observe, analyze, and improve the quality of our results even further — a great testament to the fact that research and in-depth data analysis does fuel the future of data-driven innovation that will help businesses run better.


Sophia Stolze is integrated communications manager for Customer Success Services at SAP.