Intelligent system with machine learning for the management of university pre-professional internships

Abstract

The purpose of this study was to design and perform a preliminary validation of an intelligent system based on machine learning, aimed at improving the management of pre-professional internships at a Peruvian public university. An applied research study was conducted, employing a quantitative approach and a pre-experimental design with a propositional component. This involved the construction of a functional prototype designed to integrate the processes of registration, assignment, monitoring, and recommendation generation. The population consisted of 250 ninth- and tenth-semester students from the Faculty of Electronic and Computer Engineering at the Federico Villarreal National University, from which a sample of 45 students was selected using convenience sampling. Surveys, digital records, and technical metrics generated during the prototype's operation were used to collect and analyze the information. Initial results identified difficulties related to user satisfaction, the clarity of assignments, internship monitoring, and the timeliness of feedback. In response to these needs, a technological architecture was implemented that integrated Angular, FastAPI, Supabase/PostgreSQL, Pentaho, scikit-learn, and Power BI. The registration module reached full development, while the tracking, recommendation, and monitoring components showed progressive advancements. Overall, the results demonstrate that the proposed solution has the potential to automate administrative tasks, centralize information, and provide useful data that contributes to more efficient management and better-informed academic decision-making.