The complexity of current engineered systems has increased drastically over the last decades. Due to this complexity, these systems are typically developed in collaborative settings with stakeholders from different domains involved. A pertinent example is engineering of cyber-physical systems (CPS). Such collaborative endeavours are severely hindered by inconsistencies that arise due to semantic overlaps between different models. Additionally, since the involved domain-specific languages of stakeholders may be very disparate, inconsistencies often do not manifest on the linguistic level. To cope with this problem, we propose an approach that enables better understanding how inconsistencies arise, evolve and how they should be managed. The core of our approach is a rich process modeling formalism that allows modeling multiple aspects of the development workflow in accordance with the guidelines of multiparadigm modeling (MPM). We support our approach with an open-source prototype tool for designing engineering processes, defining inconsistency patterns and their respective management alternatives, and with the ability to optimize the original process for various optimality criteria, such as consistency and costs.