Researchers develop Retrieval-Augmented Generation virtual assistant to help Maastricht University students navigate project-specific regulations

A virtual assistant based on Retrieval-Augmented Generation (a system that combines AI language models with retrieved domain-specific information) was designed to support students at Maastricht University in understanding project regulations.

The system addresses challenges such as hallucinations (false information generated by AI), missing information, and difficulty providing accurate context-specific responses in specialized domains by integrating up-to-date, domain-specific knowledge.

Through evaluation and real-life testing, the virtual assistant demonstrated effectiveness in meeting student needs while applying Large Language Models to a specialized educational context.