Job Specifications
About Luma AI
Luma's mission is to build multimodal AI to expand human imagination and capabilities. We believe that multimodality is critical for intelligence. To go beyond language models and build more aware, capable, and useful systems, the next step function change will come from vision. So, we are working on training and scaling up multimodal foundation models for systems that can see and understand, show and explain, and eventually interact with our world to effect change.
About The Role
At Luma, the Applied Research team brings our most advanced generative models to life. We sit at the intersection of Research and Product - bridging raw innovation with real creative tools used by millions. Our mission is to make Luma's multimodal foundation models more expressive, controllable, and reliable, turning frontier research into magical, production-ready experiences.
What You'll Do
Develop and maintain model variants purpose-built for specific product features and partner applications - adapting architectures, datasets, and fine-tuning strategies.
Drive continual improvements to Luma's core model-powered experiences, leading iterations that push quality, reliability, and creative depth across versions.
Collaborate closely with Product, Research, and Design to translate creative intent and user feedback into refined model behavior, intuitive controls, and new capabilities.
Build internal tools and workflows that accelerate model iteration and evaluation - enabling faster experimentation, deeper insight, and tighter feedback loops.
Contribute to applied research in safety, authenticity, and control - spanning topics like moderation, watermarking, fairness, and color science.
Qualifications
Strong engineering skills in Python and deep learning frameworks (preferably PyTorch); comfortable moving between research prototypes and production systems.
Hands-on experience with modern visual generative models (diffusion, transformers, or related architectures).
Demonstrated ability to tune, refine, and deploy models in real products using human feedback and creative evaluation.
Curiosity and passion for multimodal AI - understanding how models perceive, generate, and evolve in the wild.
Familiarity with large-scale training infrastructure (e.g., SLURM, Ray, or Kubernetes) is a plus.