Senior Pretraining Engineer (Video & Multimodal AI)

Location: Onsite in the Bay Area, California, United States; the R&D operation will be based in San Jose.

Salary: $200,000–$400,000 per year.

Company

Early-stage robotics company building an integrated AI and robotics stack for highly autonomous manufacturing.

What you will do

• Own large-scale video and multimodal pre-training runs across distributed GPU infrastructure.
• Develop and improve generative approaches, including flow matching and diffusion.
• Diagnose instability, convergence issues, and large-scale training failures.
• Improve training efficiency, reliability, and experiment velocity, and build safeguards from failed or expensive runs.
• Collaborate with robotics, simulation, perception, and systems engineers.
• Move research ideas into working systems for robotic intelligence and autonomous manufacturing.

Requirements

• Experience with large-scale video, vision, or multimodal pre-training.
• Experience with distributed training across substantial GPU compute and training large models on large datasets.
• Experience debugging difficult training failures at scale and understanding interactions between models, data, and compute.
• Experience with flow matching, diffusion models, or adjacent generative-model approaches.
• Strong software engineering and training-systems fundamentals.
• Experience designing experiments where compute cost makes mistakes consequential.
• Ability to explain unsuccessful training runs, including what failed, how the issue was diagnosed, its cost, and subsequent changes.

Culture & Benefits

• Onsite engineering environment in the Bay Area.
• Work across robot learning, reinforcement learning, multimodal foundation models, real-world perception, physics simulation, rendering, GPU optimization, and hardware-software co-design.
• Medical, vision, and dental insurance.
• 401(k).