The development of multi-tier architecture based on knowledge and model driven for the competitiveness is discussed in Higher Education; the engine are Mathematical and Computational Models for dynamic and structural analysis of the academic governance given the student high dropout. The problem is characterized by the high social and economic costs; the layers mentioned, mainly are: Institution issues, Users (3-student levels, Faculty Teachers and Parents) and the data and knowledge, enabling management of multiple piece inter-connected environments, focusing on the governance with its sustainability impact into quality, productivity and competitiveness. Data architecture gives several figures on desertion, costs (families, institutions). The period of observation, the student's retention figures were increased significantly and the desertion control was also improved.