Uma arquitetura para um sistema de análise acadêmica institucional: geração reprodutível e versionamento de indicadores acadêmicos

Abstract

The growing volume of data generated by higher education institutions has driven the adoption of Academic Analytics solutions to support academic management. However, traditional approaches often neglect fundamental aspects such as calculation reproducibility, systematic indicator versioning, and evolutionary maintainability, compromising the reliability and traceability of analyses. To address this problem, this work proposes, develops, and implements a modular architecture for institutional Academic Analytics focused on reproducible generation and versioning of academic indicators. The results demonstrate that the proposed architecture meets modularity, traceability, and reproducibility requirements, allowing distinct versions of the same indicator to coexist and enabling controlled system evolution. It is concluded that the solution provides a sustainable foundation for building institutional analytical capacity, bridging technological innovation and operational university needs, with prospects for improvement in security, scalability, and expansion of the indicator portfolio.