About this role
Train and evaluate robot policies using imitation learning, diffusion policies, and reinforcement learning fine-tuning on both simulated and real-world hardware platforms. This hands-on, research-driven position sits inside the Abaka Robotics Lab in Mountain View, California, where the team builds the data foundation for physical AI by capturing, synthesizing, and curating robotic experience.
The incoming engineer will work directly with 6-DoF collaborative robot arms and ALOHA-style bimanual platforms to test policies, run evaluations, and design scalable data pipelines that transform raw sensor inputs into high-quality training datasets. Research findings will directly shape commercial datasets for industry leaders and contribute to publications at top-tier robotics and machine learning venues.
Ideal candidates bring a solid academic background combined with practical experience handling real robot hardware and a rigorous approach to evaluation.
What you will do
- Train and evaluate robot policies across simulation environments and real-world robotic systems.
- Analyze evaluation results to determine the specific impact of different data types on policy performance.
- Reproduce relevant research papers and develop findings suitable for submission to top robotics and machine learning venues.
- Maintain training and evaluation infrastructure, including harnesses and simulation environments.
- Design scalable data pipelines that convert raw captures into training data with minimal human labeling.
What you need
- Hold an MS or PhD degree in Robotics, Computer Science, Machine Learning, or a closely related discipline.
- Demonstrate one to three years of hands-on experience working with real robot hardware through research, internships, or projects.
- Possess strong machine learning fundamentals covering optimization, generalization, and evaluation methodology.
- Understand core robotics concepts including coordinate frames, kinematics, and basic control.
- Demonstrate solid 3D vision fundamentals such as camera models, multi-view geometry, and point clouds.
- Apply rigorous evaluation standards where all claims regarding policies or datasets are backed by clear measurements.
Nice to have
- Publications or workshop papers at top-tier robotics, vision, or machine learning venues
- Experience with imitation learning or VLA codebases such as LeRobot, ACT, Diffusion Policy, OpenVLA, or π0
- Practical 3D reconstruction experience including COLMAP, SLAM, or 3DGS/NeRF
- Familiarity with simulators such as MuJoCo, Isaac Sim/Lab, Genesis, or SAPIEN
- Experience with teleoperation or data collection systems in the ALOHA, UMI, or hand-tracking lineage
Benefits
- Health insurance
- Dental insurance
- Vision insurance
- Paid time off
- Flexible work schedule
- Equity
Skills and tech stack
Questions candidates ask
What is the location and work schedule for this role?
This position is based at company headquarters in Mountain View, CA. The role is eligible for a flexible work schedule.
What is the salary range for the Robotics Research Engineer?
The base salary range for this position is $100,000 to $140,000 USD annually. Compensation may vary depending on qualifications, skills, competencies, and experience.
What hardware will I be working with?
You will work with hardware including 6-DoF collaborative robot arms and an ALOHA-style bimanual platform with full teleoperation.
What additional compensation is provided alongside base pay?
Base pay is one part of the total package, and this role is eligible for equity as well as a comprehensive benefits package.