Embodied AI Research Scientist (Humanoid Robotics)
| Hours | Full-time, Part-time |
|---|---|
| Location | Santa Clara, CA 95053 Santa Clara, California open_in_new |
About this job
Job Description
Embodied AI Research Scientist โ Humanoid Robotics
\n๐ San Francisco Bay Area, CA ยท Full-time ยท On-site
\nAbout Theta Intelligence
\nTheta Intelligence is building next-generation humanoid robots powered by advanced embodied intelligence and foundation models. Our mission: robots that understand, reason, and act autonomously in complex real-world environments
\nThe Role
\nWe're looking for a highly motivated Embodied AI Research Scientist to push the frontier of Vision-Language-Action (VLA) Models, World Models, and foundation models for humanoid robotics. You'll work at the intersection of robotics, multimodal AI, and large-scale learning systems โ building the end-to-end intelligence that lets robots perceive, reason, and act in the physical work
\nWhat You'll do
\n- \n
- Research, develop, and optimize embodied AI models โ including Vision-Language Models (VLMs), Vision-Language-Action (VLA) Models, and World Models. \n
- Build end-to-end learning systems connecting perception, reasoning, planning, and action. \n
- Develop foundation models that improve robot autonomy, adaptability, and task execution in real-world environments. \n
- Partner with robotics teams to integrate AI models with perception, navigation, manipulation, and motion-control systems. \n
- Shape technical strategy and drive key research initiatives in embodied intelligence and robot foundation modles. \n
What We're Looking For
\n- \n
- MS or PhD in Computer Science, AI, Robotics, Machine Learning, or a related fields \n
- Deep expertise in one or more of: VLMs, VLA Models, World Models, Reinforcement Learning, Embodied AI, Robot Learning. \n
- Hands-on experience developing large-scale AI models, multimodal learning systems, or robotic learning frameworks. \n
- Strong proficiency in Python and PyTorch. \n
Nice to Have
\n- \n
- Publications at top-tier venues (NeurIPS, ICLR, CVPR, ICCV, ICRA, \n
- CoRL)Experience with Robot Foundation Models, VLA systems, or World Models Track record deploying AI models on physical robotic platforms. \n
- Strong research portfolio โ publications, patents, open-source contributions, or high-impact projects. \n
Why Theta Intelligence
\n- \n
- Work on frontier research in embodied intelligence and humanoid robotics. \n
- Build foundation models that directly power real-world robotic systems. \n
- Collaborate with a world-class multidisciplinary team across AI, robotics, and systems engineering. \n
- Help define the future of general-purpose intelligent robots. \n