Software Producing Organizations are increasingly engineering Very Large Software Systems for enterprises. This presents many challenges such as service compositions, maintaining end-user variability, deployment on heterogeneous cloud ecosystems, system security, error handling and re-configuration. The AMUSE project departs from the philosophy that an enterprise model must be technology agnostic, and that there are always two transformations: (1) from the enterprise model to a technology agnostic solution model and (2) from the solution model to a technical implementation. 

Developing a technology independent enterprise model poses new challenges, especially because we are in particular developing a functional enteprise model. The enterprise model requires minimal information and is enriched with a bulk of knowledge added to any model that stems directly from decades of enterprise software development. The knowledge comes in the form of business logic from a large enterprise software domain model. 

Transforming a technical solution model to code provides many novel opportunities. To begin with, we can create a responsive architecture, that changes based on the properties of the solution model. The solution model can be further enriched with usage patterns of the software, enabling further optimization of the architecture. Currently we are uncovering new patterns for enabling the transformation, such as the CQRS pattern. Furthermore, we are developing and evaluating decision models that support the technology selection process for software producing organizations.

The AMUSE research project is an academic collaboration between Universiteit Utrecht, Vrije Universiteit Amsterdam, and AFAS Software to address software composition, configuration, deployment and monitoring challenges on heterogeneous cloud ecosystems through ontological enterprise modelling.