This study presents the development and implementation of a personalized Generative Artificial Intelligence (GAI) model, based on the ChatGPT architecture, oriented to technical teaching in the field of road intersection design. The tool has been specifically adapted for the subject ‘Intersection Design’ of the Master's Degree in Civil Engineering of the University of Alicante, through its training with current technical standards (Standard 3.1-IC, Road Junction Guide, among others). The model allows the structured introduction of geometric and traffic parameters by the user and performs analytical calculations in accordance with the established normative criteria, providing results related to crossing distances, lengths of acceleration and deceleration lanes, turning radii, and other road design elements. Validation has been carried out through practical applications in the classroom, comparing results generated with manual solutions and specialized software. Findings indicate some positive aspects in efficiency and preliminary guidance, alongside limitations in technical accuracy and inability to produce valid graphic outputs. It is concluded that, under teaching supervision, the GAI model can support the delivery of regulated content in civil engineering, but it cannot replace traditional methods or expert analysis.