Black Forest Labs

This role focuses on the post-training pipeline for multimodal generative models, including data strategy, reward modeling, preference optimization, distillation, and safety tuning. The goal is to improve model quality and align them with human intent, with a strong emphasis on shipping these improvements to users.

Research role focused on training and fine-tuning large-scale diffusion models for image and video generation, involving rigorous experimentation, ablation studies, and understanding speed-quality tradeoffs in production settings.