Research Scientist, Robotics & World Models
$160k - $185k • Remote - United States
Posted 23d ago
Job Location
Remote - United States
Tech Stack
Remote Work Policy
On-site
Categories
AI Research Engineer
About the job
Innodata is seeking a Research Scientist to lead the science of data for robotics foundation models. This role involves partnering directly with customers and frontier labs to define how data is designed, structured, and evaluated for next-generation AI systems in robotics. You will leverage scientific judgment to make high-leverage decisions about data capture, synthesis, representation, and evaluation, ensuring that the data collected and generated directly supports the development of robust and transferable AI models for manipulation, locomotion, vision-language-action, and world models.
Responsibilities
- Define data specifications for robotics foundation models, including modalities, action representations, sampling, annotation schemas, and evaluation criteria.
- Determine the optimal balance between real-world data capture and simulation-generated data, curating and weighting training mixes across diverse robot datasets.
- Guide data collection across various modalities (motion capture, egocentric, exocentric, teleoperation, multi-sensor) and synthetic pipelines to align with model requirements.
- Develop evaluation and benchmarking methodologies to predict real-world transfer, including world-model evaluation and sim-to-real gap closure techniques.
- Conduct experiments to fine-tune and evaluate foundation models using Innodata's data, focusing on data quality's impact on performance.
Requirements
- Experience in defining data specifications for robotics foundation models.
- Expertise in balancing real-world data capture and simulation for training data.
- Experience guiding data collection across multiple modalities and synthetic pipelines.
- Proven ability to build evaluation and benchmarking methodologies for real-world transfer.
- Experience fine-tuning and evaluating foundation models.