This presentation will describe how different Python tools are integrated to enable the full characterization of newly developed composite materials with reduced physical testing. Keras and Tensorflow are used to build the neural networks that function as metamodels of the FEA simulations, while emcee is used to perform Bayesian inference. In addition, core packages such as SciPy and Numpy are used to aid in the automated finite element model generation, while Surprise is used for collaborative filtering tasks David Najera, Doug Melville, Eric Jayson and Tommy Board