When dealing with processes, engineering systems need to be able to deal with the existence of many versions of the same process, known as variants. Each process variant aims to represent a specific business context but, due to some differences in the activities, resources, and control flow, they may differ in their logic. As result, the concept of a customizable process model has emerged. A process model can be customized by representing the process family in one single model deriving a variant through transformations in this single model. However, the process customization is not a trivial‐task. It must be ensured that the variant is correct. Besides, the resulting variant must respect all requirements related to the application context and internal and external regulations, among others. In addition, recommendations and guidance should be provided during the customization. In this context, this research aims to propose a framework for customizing variants according to the user's requirements. The customization is achieved by reasoning in an ontology based on the rules for selecting a variant, in the internal/external regulations, and on expert knowledge.