Propuesta de un Entorno Multimodelo de Madurez para Software Libre y su Aplicación a Plataformas de Aprendizaje Automático

Maturity models allow software to be assessed to determine whether it meets quality requirements. This paper proposes the development of a multi-model maturity framework for assessing the quality of open source software, integrating best practices from various maturity models. The presented multi-model seeks to offer a flexible solution for assessing open source software, adapting to the specific needs of an organization through the use of multiple models. In addition, this approach was applied to machine learning platforms such as TensorFlow, PyTorch, Keras, and others, in order to validate their effectiveness in this area. The results showed that the platforms evaluated present different levels of maturity, which provides key information for software selection and adoption in academic and business environments. The research lays the groundwork for future applications of this multi-model in other areas, such as machine learning service infrastructure (MLOps).