This thesis proposes an AI-driven mentor to support structured reflection processes in the context of software engineering education. The AI mentor leverages Large Language Models to help students set personalized goals and conducts weekly reflective sessions in project-based learning environments. These sessions focus on monitoring task status, analyzing impediments, and defining realistic future steps to improve project results.
The AI mentor applies reflective practices to guide students to better understand their current progress stage and ways to overcome obstacles, while focusing on identified goals. Additionally, it provides individualized, data-driven insights by integrating past reflections and connecting with project development tools, helping students stay focused and reach their full potential. This approach aims to enhance students’ self-awareness, improve their learning outcomes, and better prepare them for industry challenges.