The NephroCAGE consortium applies the latest advances of learning systems to address a multi-national healthcare challenge in nephrology. We aim to combine medical and technical innovations to build and evaluate a real-world demonstrator incorporating the expertise of two leading nations in the fields: Canada and Germany. Combining clinical data from both nations through a secure federated learning platform enables for the first time access to a unique multi-national pool of clinical nephrology data. This clinical data pool forms the foundation for applying selected Machine Learning (ML) methods to train models, which help to predict the probability for selected clinical outcomes in advance. However, the use of ML methods requires access to clinical data, which are highly protected through data protection regulation. Especially multinational projebts are hindered due to individual data protections regulations and data formats. Therefore, we propose the use of a federated learning infrastructure, where data resides at their original locations and instead of data ML algorithms and aggregated data are only exchanged. ML algorithms are executed on local data pools and gained insights from local data analysis are added to statistical models, which are exchanged between partners sites iterative learning. If the consortium outcomes prove to be successful, the developed platform and methods will be applicable to address additional medical indications and chronic diseases. Furthermore, the results would be applicable to other multinational projects, e.g. industry cooperation in automotive construction, transportation, or renewable energies.