The funding for these positions is for *basic research* on AutoML -- pure methods development without the need for any particular application. Our focus is on automated deep learning, neural architecture search, efficient hyperparameter optimization, learning to learn, and AutoML in general, and we have a track record of developing widely-used open-source tools, such as Auto-sklearn (
https://github.com/automl/auto-sklearn).