Embodied AI Engineer

Role We’re hiring an Embodied AI Engineer to build high-success, fast, and safe Physical Intelligence that works on real robots in real sites. Starting with near-term manipulation tasks, and evolving toward reusable skill libraries powered by state-of-the-art foundation models imitation learning simulation/digital-twin RL . If you want to work on embodied AI that actually leaves the lab and runs at scale, this is it. What you will own 1) Real-world manipulation skills Build manipulation behavior that can be used in real-world scenarios. This includes tasks such as grasping objects, moving them around, and placing them in specific locations. 2) Foundation models foundation models are pre-trained AI models that can be fine-tuned for specific tasks. They are the building blocks of many AI systems. In this role, you will work with state-of-the-art foundation models to build reusable skill libraries. 3) Imitation learning imitation learning is a type of machine learning where an AI model learns by imitating the behavior of a human or another AI model. This involves observing and mimicking the actions of the teacher model. In this role, you will use imitation learning to train AI models that can perform complex tasks. 4) Simulation/digital-twin RL simulation/digital-twin RL is a type of reinforcement learning where an AI model learns by interacting with a simulated environment or digital twin. This involves making decisions and receiving feedback in the form of rewards or penalties. In this role, you will use simulation/digital-twin RL to train AI models that can perform complex tasks.