Member of Technical Staff - Pretraining (AI)

TL;DR

Member of Technical Staff - Pretraining (AI): Lead large-scale pretraining experiments for multimodal foundation models (image, video, audio) with an accent on architecture, objective functions, and scaling strategies. Focus on developing novel ideas across optimizers, training algorithms, and full-stack optimizations from low-level GPU systems to high-level model design.

Location: Hybrid in Freiburg (Germany) or San Francisco (US) — at least 2 days/week in office (or one full week every other week), or remote with monthly in-person week (travel covered)

Company

Team behind Latent Diffusion, Stable Diffusion, and FLUX — foundational generative AI technologies powering tools used by millions of creators, developers, and businesses worldwide. Headquartered in Freiburg, Germany, with growing presence in San Francisco.

What you will do

• Lead large-scale pretraining experiments for multimodal foundation models, including architecture, objectives, and scaling.
• Develop and evaluate novel ideas in architecture, optimizers, and training algorithms.
• Contribute across the stack: GPU/systems optimizations, research code, and high-level model design.
• Lead independent research projects and drive cross-team initiatives.

Requirements

• Hybrid presence required in Freiburg or San Francisco — discuss specifics in interviews.
• Led or co-owned pretraining for a production foundation model (image, video, LLM, or multimodal).
• Owned architectural decisions like attention patterns, modulation, loss formulations, tokenization.
• Deep experience with large-scale distributed training (FSDP/TP/PP, 500+ GPUs, debugging scale issues).
• Track record of shipping: top publications (NeurIPS, ICML, etc.) with production impact.
• Deep Python/PyTorch proficiency; familiarity with visual generative models.

Culture & Benefits

• Distributed team with real offices; structured in-person time (2 days/week or monthly weeks, travel covered).
• Four core values: Obsessed (frontier research excellence), Low Ego (best idea wins, shared credit), Bold (ship ambitious bets), Kind (people-first warmth).
• Focus on research excellence, open science, and expanding human creativity.

Hiring process

• Discuss role specifics including location setup during interviews.