A new generation of AI is emerging in Hagenberg—one that not only performs calculations but can also prove its results. The Research Institute for Symbolic Computation (RISC) at JKU Linz and the spin-off company RISC Software GmbH are steadily expanding their role at the intersection of mathematics and AI.
From left to right: Wolfgang Freiseisen (CEO, RISC Software GmbH), Bruno Buchberger (RISC), Georg Regensburger (RISC), Carsten Schneider (Director, RISC)
AI is rapidly transforming industry, medicine, and research. However, when it is used in safety-critical systems, machine learning (ML) alone is not sufficient: results must be traceable and mathematically sound. While machine learning derives patterns from data, symbolic computation (SC)—using logic, computer algebra, and formal methods—ensures precise verification. Together, they complement each other to form a hybrid, explainable AI. RISC pursues this goal on three levels.
“AI can only become truly intelligent if it combines the statistical creativity of machine learning with the logical rigor of symbolic mathematics.”
RISC was founded in 1987 by Bruno Buchberger; today, the institute is one of the world’s leading centers for symbolic computation. The scope of the research is illustrated by two newly appointed professors: Carsten Schneider applies symbolic algorithms in particle physics—at CERN and in collaboration with DESY—while Georg Regensburger explores the structure of dynamic systems in systems biology. The SCML conference, which brings researchers from around the world to Hagenberg in July, demonstrates just how vibrant the field is. The InProSSA application project, funded by the FFG, combines machine learning and symbolic computation in such a way that the system independently selects the best solution.
To ensure this commitment to innovation in the long term, the institute is establishing the RISC European College—a two-year, English-language master’s program that combines symbolic computation and machine learning as a mathematical specialization. Funded by the State of Upper Austria and regional companies, the program uses scholarships to train professionals who will not only apply but also shape the AI of the future.
For over 30 years, RISC Software GmbH—founded in 1992 as a spin-off from the institute—has been transforming the institute’s ideas into robust solutions, ranging from safety-critical software to industrial AI and medical informatics. TaxoLogic demonstrates how this works in practice: The project brings hybrid AI to tax law, where large language models interpret complex legal sources and symbolic logic ensures transparent, legally sound results.
Together, the institute and the GmbH form a unique ecosystem of research, teaching, and application, with more than 130 experts. Since 1985, RISC has shaped the region—from the development of the Hagenberg Software Park to the creation of AI that companies can not only use but also trust.
Hagenberg Castle
Founder of RISC (Research Institute for Symbolic Computation) and RISC Software GmbH
Director of the RISC (Research Institute for Symbolic Computation)
