Case Study: Federal Employment Agency achieves better quality assurance with Capgemini’s machine learning-based code analysis

Successful Code Analysis Using Machine Learning at the Federal Employment Agency (FEA)

The Federal Employment Agency (FEA), a major German public sector organization, faced the challenge of identifying hard-to-find programming errors in its business-critical software, which processes millions of transactions and disburses billions of Euros in unemployment benefits. To avoid production issues and ensure punctual, correct payments, the FEA partnered with Capgemini to enhance its quality assurance process.

Capgemini developed a machine learning-based, static code analysis tool for the FEA. This innovative solution identifies complex code patterns and software errors by learning from verified source code, even without pre-defined manual rules. The results included optimized quality assurance, the avoidance of emergency hotfixes, and the identification of new rules affecting both functional and non-functional requirements, significantly improving the reliability of their vital software.

Federal Employment Agency

Thomas Paal

Business Unit Leader