Since February 2025, our “Data Science” workgroup has been involved in the new research project "Learn2RAG ". The project is all about "Retrieval-Augmented Generation", or RAG. RAG is an AI technique that aims to improve the performance of generative language models through a combination of retrieval and generation models. "Learn2RAG" is working on automating the learning of company-specific RAG pipelines and their use.
In addition to Paderborn University, the project consortium also includes the Institut für Digitale Technologien gGmbH (IfDT), USU GmbH, the German Red Cross Rhineland-Palatinate Regional Association and Fraunhofer IEM.
The project will run from February 2025 to January 2028 and is sponsored by the BMWK with a total amount of €2.8 million.
More information about the project can be found here.