Senior Humanoid Manipulation Research Scientist

FieldAI; Boston (MA); On-site; USD 150,000 - 300,000; Full time

Job summary

FieldAI is advancing humanoid robotics and embodied intelligence from its Boston base. You will lead research in large-scale, learned humanoid manipulation and cross-embodiment transfer, translating breakthroughs into scalable, reliable robot capabilities.

You will design rigorous experiments, validate results on real humanoid robots, and help shape FieldAI’s robotics foundation models across perception, reasoning, and control. Strong collaboration with engineers and fields teams is expected.

Qualifications

• PhD in Robotics, CS, ML, EE, ME, or closely related field.
• Strong research record in robot learning, manipulation or related areas.
• Ability to lead research projects from question to validation.
• Deep expertise in foundation models, VLAs, RL, imitation learning, or multimodal models.
• Experience training and evaluating modern deep-learning models using PyTorch.
• Experience deploying methods on physical robotic systems.
• A track record of publications or impact in robotics/ML venues.

Responsibilities

• Lead high-impact humanoid manipulation research projects.
• Develop novel model architectures and training methods for robot learning.
• Establish baselines, evaluation protocols, and benchmarks for manipulation.
• Collaborate with research engineers, systems teams, and product teams.
• Translate research into scalable training pipelines and robot capabilities.

Skills

Research leadership

Robot learning

Multimodal learning

Reinforcement learning

Imitation learning

Communication across teams

Education

PhD in Robotics/CS/ML/EE/ME

Tools

PyTorch

Job description

FieldAI is advancing humanoid robotics and embodied intelligence from its Boston base. You will lead research in large-scale, learned humanoid manipulation and cross-embodiment transfer, translating breakthroughs into scalable, reliable robot capabilities.

You will design rigorous experiments, validate results on real humanoid robots, and help shape FieldAI’s robotics foundation models across perception, reasoning, and control. Strong collaboration with engineers and fields teams is expected.