Projektsteckbrief AI-Powered Personalized Assignment System

The project aims to develop an AI-powered assignment system that tailors exercises to each student's academic background, research interests, and thesis direction. In addition, the system provides an automated preliminary evaluation of submissions, supporting instructors during review and enabling timely, detailed feedback. The solution is designed for long-term integration into Heidelberg University's Moodle platform, making personalized teaching sustainable and scalable.

The project develops an AI-based system that tailors assignments to each student's academic background, research interests, and thesis direction and provides an automated preliminary evaluation. This approach engages students in heterogeneous master's cohorts individually, eases the instructor workload, and prepares sustainable integration into Moodle.

The framework is transferable well beyond computer science to STEM subjects and any program where students have heterogeneous prior knowledge or an early thesis orientation. Prompt templates, evaluation rubrics, and a best-practice guide will be released as open materials, enabling instructors in other fields to adapt the system for scalable, personalized teaching.