This list contains only the countries for which job offers have been published in the selected language (e.g., in the French version, only job offers written in French are displayed, and in the English version, only those in English).
This is not a standard data management role; it’s a rare opportunity to be at the absolute frontier, shaping the data that trains our foundational models. You'll be the architect behind the datasets that power our generative video technology, working in close concert with our world-class research team to fuel the breakthrough models that millions of creators use. Think of yourself as the bridge between raw information and AI brilliance—you'll build the curation strategies, annotation workflows, and quality systems that help our models see and understand the world better.
Job Responsibility:
Translate researcher needs into actionable data annotation and curation strategies for our SOTA vision, 3D, and audio models
own and manage end-to-end data pipelines and annotation workflows, collaborating with external partners and labeling teams to ensure the highest quality data
provide horizontal management across multiple data pipelines, ensuring consistency and quality as we expand into new modalities
develop innovative data curation strategies, working with a diverse mix of human-annotated, raw, and synthetic data to solve complex model challenges
partner directly with researchers to diagnose model performance issues and propose data-driven solutions to improve results
define the standards for data quality and annotation excellence, establishing the foundation for how Luma scales its data operations
Requirements:
2+ years of hands-on experience in AI data operations, human data annotation, or a similar data-centric role within a top-tier AI company
direct experience translating complex researcher needs into effective data curation and annotation workflows
highly adaptable and thrive on cross-functional collaboration, with a proven ability to work across a comprehensive data pipeline, not just within a single vertical like human annotation
experience working with vision or multimodal data pipelines
a hands-on individual contributor who is driven by the work, not by people management
Nice to have:
Experience working at a company known for its SOTA vision models
experience at a top-tier data labeling provider
comfortable working with a variety of data types beyond human annotations, including raw and synthetic data
experience managing vendor relationships for data labeling and annotation
thrive in a fast-paced startup environment and have a history of taking high ownership of your work