Li-Ion batteries have been widely applied as energy storage systems, such as EVs. Data-driven methods for battery health estimation and prediction are gaining increasing interest in both academia and industry. These methods have been driven by recent advances in ML that exploit the large amounts of available data to improve BMS performance. This direction dictates the need for efficiently embedding various algorithms into a unified software framework in order to support various objectives and data requirements. In this paper, we propose an architectural framework capable of supporting several and dynamic predictive analytics processes, employing data from the heterogeneous data sources. We also present the functionalities of the framework in three scenarios in order to demonstrate its applicability.