The advances in computational methods and technology have brought materials modeling to the point where we can design new materials in silico. Even for more complicated systems like transition metal oxides, many key properties can be predicted with confidence from first principles. In this talk, two differently scoped materials design projects will be presented. The first project is a small-scale careful tuning of electronic-structure properties with composition. We have explored how pressure-induced spin transitions can be optimized in the (Fex,Mn1-x)S2 system to fit a certain application. In the second part of the talk, a high-throughput large-scale exploration of oxide materials will be presented. In this project, hundreds of CPUs and data mining screening algorithms were combined to predict new compounds with optimized target properties. An example of the potential uses of such an approach will be given.