Despite the increasing spread of digitalisation in manufacturing, humans will still play an important role in future production environments. Evidently, their role will change from physical to rather cognitive tasks, such as decision-making or control and monitoring of processes. A suitable medium that can support employees in interpreting the data generated are machine learning (ML) applications. Nevertheless, recent studies show that the knowledge required to implement an ML solution is not available in a large number of companies. In order to close the knowledge gap and subsequently prepare human operators for the implementation and use of ML applications, it is highly relevant to provide proper assistance. For this reason, the present publication aims to develop a cognitive assistance system that supports shop floor managers in implementing ML use cases in manufacturing (referred to as CAS-ML). The CAS-ML concretizes a previously published procedure model with additional steps as well as learning material and is realized as a software thereupon. Finally, the CAS-ML is evaluated by operative employees and tested on an open-source data set.