The migration of Brazilians has become more and more frequent nowadays, with the main purpose of obtaining better living conditions. Studies indicate that one of the main reasons for migration is the search for training at a high level of training. Therefore, in this scenario, this research has as main objective to analyze the exodus of Brazilian students during their academic formation process, from data extracted from their curricula registered in the Lattes Platform with the adoption of network analysis techniques. The Lattes Platform was used for referring to one of the main Brazilian academic repositories, and for having relevant information for this research. Therefore, the LattesDataXplorer framework was used for the extraction and treatment of the data. Subsequently, the data set of individuals with a doctorate completed were selected because they are individuals with a higher level of education and who maintain a constant update of their curricula. Once this was done, data was enriched with geolocation and information from the institutions where they trained, in order to obtain results from distances covered by doctors. As a way of visualizing data, network analysis was used, and metrics were used to obtain an overview of how the Brazilian scientific exodus occurs. A high concentration of doctors is perceived in cities with a higher concentration of universities that have postgraduate programs at the master's and doctoral level, as well as being characterized by having higher incomes per capita.