This paper presents an innovative software interface for the utilization of widely used machine learning algorithms in a unified Python/R programming environment. This study makes two contributions. First, a more comprehensive and specialized architecture is made available for integrating machine learning into enterprise information systems. Second, a novel software model, Rbox+, is proposed to execute machine learning algorithms by jointly leveraging the capabilities of the Python and R programming languages through an API. The proposed application programming interface (API) is tested and evaluated using a publicly available benchmark dataset for regression analysis (Car-sales dataset, available on Kaggle), applying multiple machine learning models and comparative performance metrics. The obtained results demonstrate improved computational efficiency and scalability, with the execution of multiple models completed within a short processing time on standard hardware. Unlike conventional machine learning APIs or isolated ERP analytics tools, Rbox+ enables transparent, language-independent execution and validation of machine learning models while exposing the underlying source code. The proposed approach supports practical applications in enterprise analytics, reproducible research, and machine learning education, enhancing interoperability between ERP systems, analytics platforms, and statistical programming environments.