Rhoda AI
Research Member of Technical Staff - Data & Evaluation
$200K–$300K
Mountain View · Staff · Full-time · Posted May 27, 2026
About the job
Builds large-scale video data curation, annotation, and evaluation systems that improve video model training and measure progress toward robotic tasks.
What you'll do
- Design scalable pipelines to ingest, deduplicate, filter, and classify web-scale video data.
- Develop video annotation and evaluation methods for model quality, temporal coherence, and robot-task performance.
- Use vision-language and video understanding models for automated data processing, and collaborate with training teams to guide research decisions.
What they ask for
- Strong understanding of data-centric ML and the effect of web video quality on generative models.
- Experience building large-scale video data pipelines for ingestion, filtering, deduplication, and quality scoring.
- Familiarity with video data characteristics, including temporal structure, motion quality, scene diversity, and action content.
- ML fundamentals and hands-on experience training or evaluating large generative models; ability to design diagnostic, reproducible, actionable video-generation evaluations.
Nice to have
- PhD or strong research background in ML, computer vision, or a related field.
- Experience with large-scale video dataset curation, video-generation metrics, or VLM/CLIP-style inference at scale.
Pay
USD $200K – $300K base salary
Tools
VLMs, CLIP, WebVid, HowTo100M, Ego4D, FVD, NeurIPS, ICML
AI focus
Applies video data curation and evaluation research to video foundation models and their transfer to robotic control.
Model-written summary · the original posting is authoritative