Senior Member of Technical Staff: Multimodal Pre-Training

About the Role

You'll be a senior technical leader for the pre-training of the omni-models at the core of our stack—single models trained across a wide range of objectives and modalities (LLM, VLM, video, and action) at massive scale. Working alongside our AI leadership and world-class team, you'll help set the research agenda, own high-impact architecture, objective, and data-mixture bets, and help push how far these base models scale.

Responsibilities

• Pre-Training Roadmap: Help drive the pre-training agenda for omni-models spanning language, vision-language, video, and action, in close partnership with the AI team.
• Omni-Model Objectives: Train a single model across a wide range of losses (LLM, VLM, video and image generation) at massive scale, and decide how to weight and balance them.
• Core Modeling: Lead modeling work across architecture, tokenization, and the scaling-law analysis behind compute allocation.
• Data & Infrastructure: Turn petabyte-scale text, image, and video data into training-ready mixtures with the data and infra teams.
• Technical Leadership: Help raise the bar on training reliability and evaluation, support and mentor teammates, and bring frontier ideas into our stack quickly.

Requirements

• Large-Scale Pre-Training: Deep, hands-on experience pre-training large models—LLM, VLM, multimodal, or VLA—at significant scale, with concrete results.
• Multi-Objective Training: Experience training across multiple objectives or modalities—e.g. combining language, vision-language, and generative video/image losses—and balancing them in one model.
• Generative & World Models: Hands-on experience with generative modeling of images and/or video (diffusion, autoregressive, or masked objectives) and/or world-model approaches to physical dynamics.
• Scaling & Systems: Fluency with scaling laws, distributed training, and the practical failure modes of massive runs, and the ability to design around them.
• Technical Direction & Judgment: A track record of setting direction others build on, moving fluidly between research and production-grade code, and prioritizing ruthlessly under uncertainty.

Preferred Qualifications

• Experience pre-training LLMs or VLMs at scale.
• Experience training video-generation, image-generation, or world-model systems at scale.
• Experience with action/robotics modalities or VLA models for embodied agents.
• Contributions to widely used models, influential papers, or open-source training stacks.

Benefits

• Competitive total compensation program, including salary, annual cash bonus, and company equity.
• Company-subsidized insurance programs.
• 401(k) with company match.
• Flexible PTO.
• Daily lunch.