Introduction:

The implementation of an intelligent tutor for the teaching of Stereometry in Cuba represents an opportunity to transform the learning of this discipline, overcoming the limitations of current resources. This approach would not only improve students' academic performance, but would also encourage the development of essential skills for their comprehensive training.

Objective:

Design Geotutor, an intelligent tutor system for stereometry that combines machine learning models (clustering of cognitive styles) with contextualized gamification techniques.

Methods:

The applied research was developed in three stages: 1) theoretical review and foundation; 2) diagnosis of difficulties in learning Stereometry through a knowledge test; and 3) system development. Theoretical and empirical research methods were used in its execution.

Result:

The system has a modular architecture, composed of: Student Module, Tutor Module, Domain Module and Mobile Application Interface. Geotutor operates through an initial evaluation that calibrates the system, offering personalized feedback and adaptive training to correct deficiencies.

Conclusion:

The results show that the proposal not only improves efficiency in problem solving (internal validation), but also promotes meaningful learning by integrating pedagogical strategies and technological innovation. It is concluded that its implementation in pre-university environments enhances the comprehensive training of students, aligning with contemporary educational demands.