This study analyzes how 126 undergraduate students interact with AI assistants during software engineering projects, revealing patterns and educational impacts.

• Students primarily use LLMs for concept clarification and debugging assistance
• Usage patterns evolve throughout the semester, showing adaptation to AI capabilities
• LLMs serve as personalized tutors, providing immediate feedback and guidance
• Integration of AI assistants shows promise for enhancing educational outcomes in technical fields

This research provides valuable insights for educators looking to effectively incorporate AI tools into software engineering curricula while maintaining pedagogical integrity.

Analysis of Student-LLM Interaction in a Software Engineering Project