Data Engineer, Robotics

How you will make an impact

• Build pipelines that convert public robotics datasets into CarbonSix's standardized format
• Process multimodal data including video, robot states, actions, timestamps, and sensor measurements
• Validate data quality, synchronization, and integrity throughout the conversion process
• Reliably package and upload large datasets to cloud storage
• Maintain conversion code, dataset manifests, and source provenance

What we are looking for

• Strong proficiency in Python and experience building data-processing pipelines
• Experience with Parquet, Arrow, and large structured datasets
• Familiarity with video containers, codecs, timestamps, and tools such as FFmpeg or OpenCV
• Experience with cloud object storage such as Amazon S3, Google Cloud Storage, or Azure Blob Storage
• Working knowledge of Linux, Git, and Docker
• Careful engineering habits around reproducibility, validation, and failure recovery

Nice to have

• Familiarity with robotics data such as observations, actions, and trajectories
• Basic understanding of coordinate frames, spatial transforms, and camera calibration
• Experience handling synchronized sensor data such as RGB, depth, joint states, poses, and force/torque measurements
• Familiarity with robotics dataset formats such as RLDS, LeRobot, robomimic, or Open X-Embodiment
• Experience with parallel or distributed data processing