Rhoda Ai
Research Member of Technical Staff - Data & Evaluation
Mountain View. Posted 5 months ago
The role
This person will build data and evaluation foundations for a video action model. They will design scalable curation pipelines, develop annotation frameworks, and build evaluation benchmarks.
What you will do
- Design and implement scalable curation pipelines for web-scale video pretraining data
- Develop video-specific annotation frameworks and quality filters
- Build evaluation frameworks and benchmarks to measure causal video model capabilities
- Research and implement data selection, mixing, and weighting strategies
- Deploy and scale vision-language models and video understanding models for automated annotation
- Collaborate with pre-training and post-training teams
- Track model capability trends across training runs
Required qualifications
- Strong understanding of data-centric ML and how web video data quality affects large generative model performance
- Experience building large-scale video data pipelines
- Familiarity with video-specific data characteristics
- Solid ML fundamentals with hands-on experience training or evaluating large generative models
- Ability to design evaluations for video generation models that are diagnostic, reproducible, and actionable
Nice to have
- PhD or strong research background in ML, computer vision, or a related field
- Experience with large-scale web video dataset curation (e.g., WebVid, HowTo100M, Ego4D, or similar)
- Familiarity with video generation quality metrics (FVD, perceptual quality, motion consistency)
- Experience running VLM or CLIP-style inference at scale for automated video filtering and annotation
- Prior work on evaluation methodology for video generation or world models
- Understanding of how web video data properties connect to downstream robotic action prediction
- Publication record at NeurIPS, ICML, ICLR, CVPR, or related venues
About the company
Rhoda AI is building the next generation of generalist intelligent robots. They own the full robotics stack from hardware to foundation world models.