Research Member of Technical Staff - Data & Evaluation

Highlights

• Own web-scale video curation shaping pretraining for generalist robots.
• Build benchmarks measuring causal video prediction and rollout fidelity.
• Drive data quality that compounds across every training run.

What they offer

• equity

Ideal candidate

A data-centric ML researcher who obsesses over data quality and honest evaluation. You enjoy building large-scale systems that accelerate research and translate video understanding into real robot capability.

Requirements

• Strong understanding of data-centric ML and impact of web video quality on generative models
• Experience building large-scale video data pipelines for ingestion, filtering and deduplication
• Familiarity with video data traits like temporal structure, motion quality and action content
• Hands-on experience training or evaluating large generative models with solid ML fundamentals
• Ability to design diagnostic, reproducible evaluations for video generation models