Crowd work is a popular form of digital work. Anyone can work on crowd work platforms and it is a great way for companies to outsource tasks externally. This new form of work has developed strongly over the last few years, so that crowd workers and politics are now faced with grave grievances. This results in three problems: 1. There is no way to communicate with each other on a crowd work platform. 2. There are too many forums and chats where crowd workers can communicate with each other. This leads to a dissemination of knowledge. 3. Members of interest groups (e.g. Chamber of Labor in Vienna, ÖGB, IG Metall) have no possibility to get in direct contact with the crowd workers. In addition, there is a lack of opportunity to regularly collect the working conditions of the crowd worker. Our goal is to develop a central solution that addresses all these problems. Based on our research questions we first identified the needs of crowd workers, platform operators, employers and other stakeholders (e.g. chamber of labor in Vienna, ÖGB, IG Metall). Then we elaborated seven design principles and an architectural system design. These are intended to highlight key features to meet the needs of crowd workers and other stakeholders and to map identified business processes. The proposed approach follows an architectural design of a collective intelligence system (CIS). A CIS allows indirect communication among the users. This type of communication animates the system and leads to a lively exchange among users. To determine the correctness and usefulness of the design principles and the architectural design, a prototype has been implemented. In a case study, stakeholders were invited to evaluate the prototype. The results show a very good match between the design principles and the CIS approach for the crowd work domain. Therefore, it is concluded that the goal of the work has been achieved through the determined design principles and system design which optimally support crowd workers and stakeholders through a central platform.