The study presents the development of a platform for structuring adaptive courses based on active, reflexive, theoretical and pragmatic learning styles using artificial intelligence techniques. To this end, the following phases were followed: search, analysis and classification of information about the process of generating content for courses; analysis and coding of the software component for generating content according to learning styles; and application of tests for validation and acceptance. The main contribution of the paper is the development of a model using neural networks and its integration in an application server to determine the contents that correspond to the active, reflexive, theoretical and pragmatic learning styles.