In the last years, as object-oriented software systems became more and more complex, the need of performing automatically reverse engineering upon these systems has increased significantly. This applies also to enterprise applications, a novel category of software systems. As it is well known, one step toward a research infrastructure accelerating the progress of reverse engineering is the creation of an intermediate representation of software systems. This paper shows why existing intermediate representations of object-oriented software are not suitable for performing reverse engineering upon enterprise applications and proposes an intermediate representation (a model) for enterprise applications which facilitates the process of reverse engineering upon this type of applications. Based on an experimental study conducted on three enterprise applications, we prove the reliability of the introduced approach, discuss its benefits and touch the issues that need to be addressed in the future.