Company Description
EmbodyX is a Silicon Valley-based startup dedicated to developing cutting-edge robotic AI and software to address the global labor shortage through innovative technologies. Our team comprises experts from top Silicon Valley tech companies such as Google and Snapchat, as well as renowned university robotics labs including Northeastern University, CMU and MIT-IBM Watson. We are committed to pushing the boundaries of technological advancement to create impactful solutions.
We accept full-time employees and interns as well but the candidates must be able to start in Mar/April 2026
Role Description
• Develop and test robotic automation pipelines using commercial robotic arms (e.g., Interbotix, WidowX).
• Implement and evaluate state-of-the-art imitation learning and policy models (e.g., ACT, OpenVLA, π0, π0-Fast) with Hugging Face frameworks such as lerobot.
• Benchmark policies on latency, generalization, and action smoothness across diverse real-world and simulated tasks.
• Build and extend task environments in simulators such as Isaac Gym and MuJoCo, incorporating domain randomization for effective sim-to-real transfer.
• Collaborate with the engineering team to deploy policies in real-world robotic systems and analyze performance in dynamic workflows (e.g., object manipulation, drawer handling, container placement).
• Design and iterate on custom robotic hardware platforms, including collaborative dual-arm setups and humanoid systems, to support advanced Physical AI research.
Qualification:
• Strong programming skills in Python and experience with robotics frameworks (e.g., ROS, Isaac Gym, MuJoCo).
• Familiarity with machine learning libraries (PyTorch, TensorFlow) and reinforcement/imitation learning techniques.
• Must have real robotic development project experience.
• Proficient with Linux environments and experienced with Nvidia Jetson and related platforms.
• Strong ability to read academic research papers and reproduce experimental results.
• Hands-on experience with robotic hardware and control pipelines is preferred.
• Optimize mechanical and electrical integration for high-performance actuators and sensors, ensuring hardware robustness for complex manipulation and locomotion.
• Completion of at least one internship or full-time role in robotics, AI, or a related engineering field is preferred.
What We Offer
• Opportunity to work on cutting-edge robotics and embodied AI projects.
• Direct experience deploying algorithms on real robotic systems.
• Mentorship from a team of researchers and engineers at the intersection of robotics, AI, and simulation.
Others
• This position may involve travel to on-site field locations for deployment and testing
• We are currently recruiting candidates who can start in Spring 2026. If you are seeking a Summer 2026 internship, please watch for future postings.
• The position is based at our headquarters in the San Francisco Bay Area, and relocation is recommended for this position.
