Staff Engineer, World Models for Embodied AI

Veeda AI; Seattle (WA); On-site; USD 150,000 - 230,000; Full time

Job summary

Veeda AI is building multimodal foundation world models for Physical AI in Seattle. Join a fast-moving team tackling challenges at the intersection of AI, robotics, and embodied intelligence to push the boundaries from day one.

You will design and train action-conditioned video and latent dynamics models, recover pseudo-actions from unlabeled data, and ensure long-horizon stability with memory representations. Strong ML experimentation and PyTorch expertise required.

Qualifications

• Bachelor's degree or equivalent hands-on experience in computer science, engineering, or a related field.
• Experience training generative or predictive sequence models end-to-end at multi-node scale (video, 3D, or latent dynamics).
• Fluency in PyTorch from prototype to production-scale training.
• Ability to independently design, execute, and analyze machine learning experiments.
• Understanding of diffusion and flow matching, long-context sequence modeling, and practical limits.

Responsibilities

• Design, train, and scale action-conditioned video and latent dynamics models.
• Condition rollouts on robot action chunks and camera trajectories; recover pseudo-actions from unlabeled video.
• Train to prevent drift by using model rollouts and persistent scene state representations.
• Fine-tune against physics-grounded reward models and verifiers; distill to few-step samplers for interactive rates.
• Build evaluation methods for action-following, physical plausibility, and long-horizon drift.

Skills

PyTorch

Generative sequence models

Experiment design

Long-context modeling

Model evaluation

Education

Bachelor's degree or equivalent hands-on experience in Computer Science or Engineering

Tools

Diffusion models

3D scene representations

Job description

Veeda AI is building multimodal foundation world models for Physical AI in Seattle. Join a fast-moving team tackling challenges at the intersection of AI, robotics, and embodied intelligence to push the boundaries from day one.

You will design and train action-conditioned video and latent dynamics models, recover pseudo-actions from unlabeled data, and ensure long-horizon stability with memory representations. Strong ML experimentation and PyTorch expertise required.