In this talk I will discuss how Scientific Workflows can be used to model a wide range of science and engineering problems, and how these are often phrased as optimisation problems. I will show that workflow engines provide an attractive programming environment for both specifying and executing the problems, and that it is relatively natural to introduce optimisation into these. This allows a user to not only model complex systems, but also ask what parameter values optimise the system using automatic search algorithms. For example, an engineer may augment a workflow that models an aircraft engine with features that allows them to determine which fan blade shapes parameters are optimal. I will also discuss how a user can be involved in the optimisation process through modern science gateways.