This practice‑based study examines the implementation of an AI‑supported tutoring system (CircleIn) in undergraduate Mathematics and Biology courses to explore its influence on student engagement. Embedded within active‑learning assignments, the AI tutor was intentionally designed as a learning scaffold rather than a shortcut, supporting problem generation, immediate feedback, concept clarification, and peer collaboration. Using a mixed‑methods approach, the study analyzes AI usage metrics, course performance data, and anonymous student surveys. Preliminary findings suggest the AI tutor supports behavioral, cognitive, and emotional engagement, particularly for quiet students, and enhances understanding of complex STEM concepts.Keywords: Engagement; AI-supported Tutor
