This paper presents the current state of the Communication Support
System (CoSy) which is a web-based system meant for assisting com-
munication training in the healthcare domain. To improve feedback
quality and thus learning outcomes of future healthcare profession-
als, CoSy generates feedback regarding a case-based student-actor
conversation. The feedback contains verbal and para verbal aspects,
i. e. speech rate, used word, and conversation share. Therefore, the
system implements three components: (1) Audio recording of the
conversation, (2) AI-based analysis of that conversation, and (3)
visualization of corresponding feedback. In line with the large-scale
project LABORATORIUM, a human-centered design approach is
used to design the system, while a machine learning algorithm is
built and trained with domain experts to guarantee reliable speech
analysis. The result is a flexible and assistive system that can be used
to complement communication training in which various health-
care situations can be practiced and discussed in class. Future work
will also face a real-time analysis as well as more complex analysis
aspects of communication like emotions or paraphrasing.