Machine Learning and AI Revolutionize ERP Testing

Artificial intelligence and machine learning can help test, test, and retest. Most organizations today rely on complex IT system architectures to manage the multitude of different tasks. The core is usually a cloud-based SAP S/4, supplemented by other specific software solutions. This provides SAP customers with powerful IT. However, this also results in numerous interfaces and media disruptions—which means many potential sources of error, for example as a result of unclean data transfer.

In order to ensure cross-system processes, organizations must regularly test their system environments. With frequent updates and upgrades, this is a labor-intensive task. Modern processes based on the use of AI (artificial intelligence) and machine learning concepts can help.

"Companies have been increasingly focusing on test automation for some time now," says Thomas Steirer, a test and AI specialist at digital engineering firm Nagarro, who is currently involved in several research projects in this area. "But they often don't use the most modern methods—partly because they simply don't have an overview of what is already technically possible."

"It is essential for consultants to gain an overview of a company's current processes in the shortest possible time frame."

Thomas Steirer,
Testing and AI specialist,
Nagarro

This includes, for example, the question of where to apply modern test structures. Thomas Steirer points out that the automatic error analysis of failed test cases in particular can provide useful insights for companies: "Our practical experience shows that many failed system tests can usually be traced back to just a few causes. Once these are known, IT departments can optimize their test infrastructure in a much more targeted manner. To do this, they need to evaluate and classify failed test log files using machine learning models. Unfortunately, this is often a "catching-up" process.

This process is not trivial. In order to achieve a suitable classification, test experts must first train the necessary ML algorithm. They give it the common failure categories and then practice the correct assignment manually using training data.

To the partner entry: