My current research goal is to develop embodied artificial intelligence capable of human-level motor control. The focus of my PhD has been on training agents which can control realistic models of the human body. We can learn a great deal about how the brain controls our body, when we can study an artificial agent able to locomote and to dexterously manipulate objects. More recently, I have developed autonomous agents for advertisement bidding. Previously, I worked on traffic optimization with Reinforcement Learning.