Event
The 48th German Conference on Artificial Intelligence, organised by the German Informatics Society, is one of Europe's most important AI conferences. That Perelyn was represented this year with not one but two workshop contributions is, for us, a sign that our research is finding fertile ground.
At the Special Interest Group for Knowledge Management workshop, Michael Banf presented an approach we have been working on intensively: Tripartite GraphRAG via Plugin Ontologies. The idea behind it is a specialised GraphRAG representation that organises knowledge structures in a way that enables AI systems to access information more precisely and context-sensitively.
In the afternoon, Johannes Kuhn took the stage at the "AI in Production" workshop. His topic: How can language-model-specific standards such as the Model Context Protocol improve communication between AI systems and manufacturing environments? The question sounds technical, but it has a very concrete core. In industrial manufacturing, standardised interfaces for AI integration are still largely missing. And that is exactly what holds many companies back.
What stood out to us: the discussions after the talks. Johannes' contribution sparked an intense exchange about the complexity of connecting disparate systems in production environments. And the knowledge management workshop led to conversations we are keen to continue.
Research and practice belong together for us. Conferences like this one show where the bridge holds.
News
The proceedings from the "AI in Production" workshop at KI2025 have been published. Among the contributions is a paper by our team on the use of the Model Context Protocol (MCP) in industrial production environments.
News
Dominik Filipiak and Michael Banf are co-authors of a community paper on the 2025 Topological Deep Learning Challenge, published in the Proceedings of Machine Learning Research. Their contributions feed into TopoBench, an open benchmarking library for the research community.
