Machine Learning Scientist 4 - Studio Media Algorithms

Research and prototype reusable algorithms for the full ML lifecycle in multimedia, including data processing, feature extraction, and GPU-accelerated training. Design, train, post-train, fine-tune, and evaluate generative computer vision models with applications across production workflows. Develop deployment strategies to bring models into production and measure impact with robust evaluation. Mentor junior scientists and define impactful research directions while collaborating with ML engineers, product managers, and creative partners. Expand technical depth into adjacent areas as business needs evolve, including 3D vision and graphics when relevant. Requires 5+ years of research experience, deep familiarity with diffusion/autoregressive models, strong DL framework skills (PyTorch/TensorFlow), and programming proficiency in Python, C++, or Java.