Research Scientist, Robotics & World Models

InnodataRobotics, Data Engineering

Remote - United StatesFull-TimeSenior

Salary160,000 - 185,000 USD per year

Apply NowOpens the employer's application page

Job Details

Roughly 4+ years of hands-on industry experience in robot learning or robotics ML.
Bachelor's degree in computer science, electrical engineering, robotics, or related technical field (Advanced degree preferred).
Strong PyTorch fundamentals and experience training/evaluating robot policies via imitation or reinforcement learning.
Experience curating, filtering, and weighting robot data across different embodiments and sensors.
Fluency in robotics data formats such as LeRobot, RLDS, and Open X-Embodiment.
Hands-on experience with simulation and synthetic data (e.g., NVIDIA Isaac Sim, MuJoCo).
Experience with teleoperation or egocentric data collection.
Experience fine-tuning large VLA models using HuggingFace transformers and PEFT.
Track record of first-author publications or open-source contributions at venues like CoRL, ICRA, IROS, RSS, or NeurIPS.
Ability to clearly explain data and modeling decisions to expert and non-expert audiences.

Translate requirements of vision-language-action models and locomotion policies into concrete data specifications.
Curate and weight training mixes across heterogeneous robot datasets and sensors.
Guide data collection across various modalities and synthetic pipelines.
Build evaluation and benchmarking methodology to predict real-world transfer and sim-to-real fidelity.
Run experiments including fine-tuning and evaluation to demonstrate model improvement.
Design adversarial and long-horizon evaluations to identify policy failure modes.
Publish benchmarks, methodology, and papers to advance the field.
Collaborate with capture labs and annotation teams to execute collection and labeling plans.

View Full Description & ApplyYou'll be redirected to the employer's site