Novel nano-architectured materials are at the heart of the next generation technologies. Most of them are composites, exhibiting complex ordering on multiple length scales. The optimal architecture can be extremely difficult to identify. The broad design space of different elemental compositions, structure phases, and microstructures is challenging to explore. Efficient discovery of the optimal solution requires an infrastructure able to combine different information sources with reliable predictive models. Different pieces of information together provide a more complete picture of the problem than independent studies. To harness the opportunity offered by engineered nanomaterials strong models have to be extracted from a limited set of configurations. Multiscale computational methods provide an efficient mean for the systematic investigation of the broad design space. Our goal is to complement experimental study for harvesting of both mechanical/kinematic models and solution optimization.