With the increasingly intense use of virtual educational environments, significant amounts of data are generated that need to be processed and studied to assist in the qualification of the educational process. In this context, the Learning Analytics process emerges, which seeks to provide insights into student behavior in different learning environments and contexts. However, due to the large amount of information contained in the data generated, the task of interpreting and illustrating this information ends up being complex. Therefore, this study seeks to demonstrate how generative artificial intelligence can help in understanding and planning the analysis of this data, in order to enable the creation of research problems and suggest appropriate visualizations for the data sets. Seeking to unite these fields, the present work used a database created over 4 years, which contains information from 8,836 students from a university in Uruguay. Thus, with the support of generative AI, a research problem was created using student grades grouped by subject, and a visualization of stacked bar graphs. As a result, a superficial analysis demonstrated evidence of significant variations in the same discipline in a broad context, in terms of academic performance, when compared with all the data, thus enabling new perspectives to improve the development of disciplines and, consequently, student performance.