ML Research Engineer, Data
Core
Architect and build data pipelines that process terabyte-scale robot fleet data (video, sensor streams) to train and evaluate models for home and business robots.
Role type
Senior ML Research Engineer (Robotics Data)
Builds
Data pipelines, training ingest systems, and evaluation benchmarks for robot learning models
Domain
Robotics, Machine Learning, Data Engineering
Deliverable
production ML models
Required skills
Data engineering at scale, Python, software engineering fundamentals, signal processing, statistics, unsupervised structure-finding, data debugging, model loss analysis, embedding search, VLM-assisted filtering
Preferred skills
C++, batch processing at scale (Ray, Spark, Dask), workflow orchestration (Airflow, Kubeflow), robotics data pitfalls, robot learning exposure (VLAs, world models, RL)
Technologies
S3, GCS, Ray, Spark, Dask, Airflow, Kubeflow, VLM
Responsibilities
Characterize data coverage, redundancy, and drift; identify and automate removal of bad data (bad trajectories, annotations, dropped frames); build data lifecycle tools (curation, preprocessing, augmentation, versioning); partner with researchers to build targeted datasets for model failures; build evaluation datasets and benchmarks
Seniority
Senior, hands-on IC