Luma AI is looking for a Research Scientist / Engineer – Data to build the fundamental data layer that unlocks advanced capabilities in our foundation models. In this role, you’ll tackle open-ended challenges around how different modalities combine to create powerful and versatile AI systems.

What You'll Do

Identify capability gaps and research solutions
Design datasets and data-mixture ablations to systematically improve model capabilities across vision, audio, and language
Develop evaluation frameworks and benchmarking approaches for multimodal AI capabilities
Create prototypes and demonstrations that showcase new multimodal capabilities

What We're Looking For

Strong programming skills in Python and PyTorch
Experience with large-scale dataset creation and management
Experience with multimodal data processing pipeline
Understanding of computer vision, audio processing, and/or natural language processing techniques

Nice to Have

Expertise working with interleaved multimodal data
Hands-on experience with Vision Language Models, Audio Language Models, or generative video models