My research focuses on the components needed for unlocking the path to robotic embodied intelligence. I am researching methods methods at the intersection of robotics and machine learning for mobile manipulation systems by enhancing model-based methods with the adaptation properties acquired by exploration and learning. This synergy will enable robots to solve long-horizon household tasks while interacting with humans and their environment, taking a step towards my vision for embodied AI assistants. Particularly important in my research is understanding how to develop algorithms by coupling perception and action. Inspired by human cognitive development, I believe that robotic skills and reasoning can emerge and develop through the active exploration for new information through observation and interaction with the world. A particular interest of mine is the investigation of structural inductive biases in learning, which can lead to learning world models of the robot and the environment asa whole, so that the robot can plan and react taking into account the environment and the perceived uncertainties as a whole.