Design and Implementation of a Personalized AI Tutor for University Education Using RAG

Abstract

This paper presents the design and implementation of an AI-based educational assistant aimed at supporting university students in their learning process. The system utilizes Retrieval- Augmented Generation (RAG) [1] to provide personalized guidance and encourage active learning through questions and exploration rather than supplying direct answers. Technical implementation focuses on a dynamic chat interface combining modern AI models [2], vector databases for retrieval, and advanced prompt engineering strategies. Preliminary observations based on simulated interactions suggest that the tutoring approach encourages deeper engagement and critical reasoning. This aligns with findings from prior research indicating that AI systems promoting active student reflection can lead to improved learning outcomes compared to direct-answer systems [6], [7]. Future improvements aim to include features such as knowledge gap identification, adaptation to the student's current understanding level, and curriculum-aware guidance.