Applied Researcher (AI)

Location: Redwood City, California, United States; on-site

Company

Develops general-purpose robots powered by an embodied AI foundation model that learns and improves across varied real-world environments.

What you will do

• Design and ship continuous learning pipelines that convert deployment successes, failures, and teleoperation corrections into targeted fine-tuning and online policy improvements.
• Analyze high-frequency multimodal sensor and video data across tens of thousands of fleet episodes to identify failure modes, drift, and regressions.
• Apply offline reinforcement learning, RL fine-tuning, and reward modeling from human and teleoperation feedback to improve deployed robot policies.
• Build real-time monitoring for anomalies, near-failures, and out-of-distribution scenes, including human-versus-automated intervention decisions.
• Characterize and reduce generalization gaps across new customer sites, lighting conditions, layouts, and objects.
• Own the path from data investigation to evaluation harnesses, production dashboards, and deployed model updates in collaboration with Research, Data, and Deployment teams.

Requirements

• Hands-on experience in at least two of reinforcement learning, sensor-data modeling or anomaly detection, vision-language models, and continual or online learning.
• Experience building monitoring, evaluation, or data pipelines for live machine learning systems and working with real-world fleet data.
• Experience designing and interpreting production experiments, including A/B tests, canary rollouts, and staged fleet deployments.
• Strong Python and PyTorch or JAX skills, with experience handling large multimodal datasets and distributed compute such as Slurm or GPU clusters.
• Bachelor's, Master's, or PhD in computer science, robotics, statistics, or a related field, or equivalent practical experience.
• Ability to turn fleet-scale data investigations into clear recommendations for researchers and operators.

Nice to have

• Experience with robot fleets or other physically deployed autonomous systems in the field.
• Experience building or fine-tuning perception or foundation models for monitoring, captioning, or anomaly detection.
• Experience with change-point detection, forecasting, or other statistical methods for sensor and telemetry anomaly or drift detection.
• Experience with human-in-the-loop learning, reward modeling from operator corrections, active learning, or failure-case data curation.

Culture & Benefits

• Emphasis on shipping measurable improvements to real robot fleets rather than optimizing benchmarks or producing research for its own sake.
• Work alongside researchers and engineers with backgrounds at leading technology companies and universities.
• Technical rigor, mutual respect, and commitment to diversity and equal opportunity.
• Opportunity to work on robots deployed across multiple commercial industries and real-world environments.