Robotics ML Research Engineer — Real-World Autonomy

Cubiq Recruitment; Palo Alto (CA); On-site; USD 140,000 - 190,000; Full time

Job summary

Cubiq Recruitment in the San Francisco area is seeking a Research Scientist to architect and deploy machine learning models powering next‑generation robotic systems. You’ll work hands‑on with real robots in a rotating team, translating cutting‑edge research into robust autonomy.

The role emphasizes imitation learning, RL, and vision-language-action models, with opportunities to lead independent research and collaborate with a design‑minded engineering group in a fast‑paced startup environment.

Qualifications

• PhD (or near completion) in Robotics, ML, or related field.
• Expertise in robot learning - imitation learning, RL, or Video Generation.
• Experience deploying models on real-world robotic systems.
• Ability to lead independent research and thrive in fast-paced environments.
• Excellent communication skills and collaborative mindset.

Responsibilities

• Train, deploy, and maintain manipulation models on physical robots.
• Build autonomy stacks using imitation learning, RL, and VLA approaches.
• Develop robust teleoperation and data collection frameworks.
• Extend state-of-the-art robotics research and adapt it for real-world use.
• Iterate hands-on with real robots in the SF lab.
• Collaborate with a highly skilled, design-conscious engineering team.

Skills

Robot learning

Imitation learning

Reinforcement Learning

Vision-language-action models

Education

PhD in Robotics or ML

Near completion

Job description

Cubiq Recruitment in the San Francisco area is seeking a Research Scientist to architect and deploy machine learning models powering next‑generation robotic systems. You’ll work hands‑on with real robots in a rotating team, translating cutting‑edge research into robust autonomy.

The role emphasizes imitation learning, RL, and vision-language-action models, with opportunities to lead independent research and collaborate with a design‑minded engineering group in a fast‑paced startup environment.