Innovative assistance system supports students at EAH Jena with personalized learning content.

On June 19, 2025, the KiWi project—an interactive, AI-supported tutor for students—was presented at the eTeach Network Thuringia dialogue workshop in the auditorium of the Ernst Abbe University of Applied Sciences in Jena.

Providing individual support to students with different learning requirements is a key challenge in higher education. KiWi is a novel assistance system based on large language models (LLMs) that provides targeted support to students.

At the heart of KiWi is the principle of retrieval-augmented generation (RAG): relevant information is retrieved directly from lecture slides and then processed by the language model to provide precise, context-related answers to student questions. This allows learning content to be personalized and knowledge gaps to be closed efficiently.

The presentation highlighted the development of the system, technical challenges, and initial experiences from test phases. In addition, prospects for future enhancements were presented—from new fields of application in teaching to the integration of additional AI technologies.

Spontaneously stepping in for the originally scheduled speaker, Richard Klein provided exciting insights into the project, which shows great potential for the further development of digital university teaching.