Model extraction and explanation are essential for interpreting the decision logic of complex systems. One example is the complex decision system that assigns Identified Actions (IAs) to applications that are planned for migration to the public cloud. In this paper, we show that it is possible to extract interpretable surrogate models that approximate the behaviour of these systems by leveraging application metadata and historical observations. Our approach enables application owners to understand and anticipate identified actions assigned to their applications, providing actionable insights for more efficient cloud migration planning. Using Random Forest models and explanation techniques such as SHAP and Bellatrex, we achieve an average F1-score of 91% for the top 20 IAs, demonstrating the effectiveness of model extraction and explanation in this context. We discuss current challenges and invite future work to further advance these methods.