SigOpt was a machine learning experimentation and optimization platform whose API applied ensemble Bayesian and black-box optimization to tune model hyperparameters, run design-of-experiments, and track experiment metadata at scale. - View it on GitHub