Facing the challenges of the digital age concerning lifelong learning, this contribution presents an approach to dynamically establish Connectivist communication networks. According the statement "the pipe is more important than the content within the pipe" by Georg Siemens, learning in digital age includes the connection of people to share required information. For this purpose, the Wiki-Learnia learning platform, which collects context information about users, is combined with the InterLect software, which identifies topics and semantic relations of contents. Based on mapping of both data sources, a wide range of matched users can be found for a specific content-related communication channel. By analyzing the course of conversation at runtime, participants can adaptively be added and removed from communication. Consequently, the presented solution serves as a just-in-time learning approach for finding direct help by experts. [For full proceedings, see ED571332.]