This paper explores the integration of exact optimisation approaches into a Manufacturing-as-a-Service (MaaS) platform. Specifically, it addresses the scheduling of production requests within a flexible and distributed real-wold manufacturing network in the plastic injection molding domain. While traditional Mixed-Integer Linear Programming (MILP) formulations struggle with scalability in realistic-sized instances, a Constraint Programming (CP) model demonstrates higher efficiency, providing near-optimal solutions within short computational times. Formally, we examine the flowshop scheduling problem under multiple objectives that reflect the diverse interests of MaaS stakeholders, i.e., the minimisation of makespan and total so-called ‘locality’ weights. To balance these conflicting objectives, Pareto frontiers are constructed to provide insights for managerial decision-making. Experimental results show that the CP model outperforms the MILP formulation in both solution quality and scalability. Moreover, the findings suggest that greater emphasis on makespan ensures better objective balance without significantly impacting locality weights. Overall, our work contributes to the ongoing discourse on MaaS operations, showcasing how exact optimisation methods can enhance scheduling efficiency in distributed manufacturing ecosystems.