The integration of IT into educational processes based on knowledge requires an interdisciplinary pedagogical–informatics approach. However, current technology cannot optimally synchronize pedagogical theories (e.g., learning styles, Bloom's taxonomy, and TPACK framework) with IT tools (e.g., information processing and software for adaptive, blended, and collaborative learning). The key problem is the absence of a universal representation of knowledge that can be understood by both humans and computers. However, both pedagogical research and technological research operate only with the abstract concept of knowledge, so teachers must adapt to the existing technologies. The authors solve this problem in university teaching and research by introducing, as a representation of knowledge, so-called virtual knowledge, which is a specific information structure in the form of a database table. Teachers and students can manually or automatically insert learning content, in the form of unstructured and heterogeneous data, into the tables. Accordingly, it is structured so that it can then be processed by software and the activities of teachers are automated using hypertext links to the hybrid environment. Using examples from university practice, the authors explain the non-relational database paradigm of creating knowledge tables and how such an all-in-one system works as personal IT support for teachers. It consists of the educational software WPad, WPad BVI (which was tested in teaching blind and visually impaired students) and PIKS, and a hybrid IT infrastructure. In the form of screenshots, the authors illustrate how the learning content was inserted into the knowledge WPad tables within adaptive, blended, and collaborative learning. They also present the theoretical idea of a CSU infinite time loop of knowledge processing, which is the basis for future research. The novelty of the presented IT system is supported by registration at the patent office.