Research Engineer, Multimodal Data
$150k - $300k • San Francisco • FullTime
Posted 5mo ago
Job Location
San Francisco
Tech Stack
Remote Work Policy
On-site
Employment Type
FullTime
Categories
AI Research Engineer
About the job
Eventual is building data curation for Physical AI, processing petabyte-scale video and sensor data to extract signals about how hands and objects interact, temporal events, and data reliability. These signals are used by leading Physical AI labs and robotics teams to select training data for their models. The work involves large-scale data systems, computer vision, and ML research, with a direct impact on frontier model training. The company is a small, well-backed team with experience from major tech companies, aiming to build the foundation for the next generation of Physical AI.
As a Research Engineer on the Visual Understanding team, you will be responsible for making petabytes of video data queryable by content. This involves defining the roadmap for visual understanding capabilities, training and selecting models for corpus-scale annotation, and building rich datasets for customer models. This is an applied research role where you will conduct experiments, ship to production, and directly influence customer models and robots.
Responsibilities
- Own the visual understanding roadmap from model selection to production inference at scale.
- Train, fine-tune, and evaluate vision-language models (VLMs), VQA models, embedding models, and CNNs.
- Reduce per-clip annotation costs through model selection, distillation, and pipeline optimization.
- Build queryable datasets for customer training, including designing taxonomies and ensuring quality.
- Collaborate with dataloading and storage teams to ensure seamless data flow.
- Work directly with researchers at partner labs to gather feedback and iterate on models.
Requirements
- Strong familiarity with modern vision and multimodal models (CNNs, VLMs, VQA, embeddings) and their deployability.
- Experience running these models at scale on real video and sensor data, preferably for perception tasks.
- Background from a perception team in self-driving, robotics, or visual-data companies, or equivalent research experience.
- Comfort with cloud infrastructure and large-scale data processing, including shipping jobs that run on thousands of GPU-hours of video.
- Experience training and evaluating vision or multimodal models.
- ML/AI research background (papers, citations, or research organization experience).
- Experience with embeddings, retrieval, or content-aware search at scale.
- Experience designing labeling taxonomies or managing annotation programs.
Benefits
- Competitive compensation
- Meaningful startup equity
- Catered lunches and dinners
- Commuter benefit
- Team-building events
- Health, vision, and dental coverage
- Flexible PTO
- Latest Apple equipment
- 401(k) plan with match