Remote Senior Multimodal Data Curation Scientist

NVIDIA; Town of Texas (WI); On-site; USD 224,000 - 357,000; Full time

Job summary

NVIDIA is seeking a Senior Applied Research Scientist to advance deep learning data curation and document extraction pipelines for foundation-model training. You will tackle multi-modal data (documents, images, audio, video) at petabyte scale, building efficient models, and contributing to the Nemotron Parse ecosystem.

You will collaborate with ML engineers and researchers to deploy NIMs, publish findings, and mentor juniors while serving as a key contributor to production-ready pipelines and

Qualifications

• Master's/PhD or equivalent in data curation, document AI, information retrieval or multimodal research.
• Publication record in CVPR, ICCV, ECCV, KDD or similar conferences.
• Experience deploying multimodal models at scale and in production environments.
• Strong Python and PyTorch hands-on experience.

Responsibilities

• Develop efficient and performant models and data pipelines for multi-modal data (documents, images, audio, video).
• Build petabyte-scale extraction and deduplication pipelines with document and HTML parsing.
• Scale pipelines across hundred-node GPU clusters for high-quality training data.
• Craft datasets, metrics, experiments, and validation scripts for research methodologies.
• Develop NVIDIA Inference Microservices and blueprints to deploy NIMs in pipelines.
• Publish papers, blogs, and training materials to help customers use research results.
• Stay current with data curation advances in academia and industry.

Skills

Python programming

PyTorch

Multimodal data processing

Documentation and papers

Communication skills

Mentoring junior engineers

Education

Master's degree or PhD in data curation / multimodal research

Tools

Ray

Spark

Dask

NVIDIA Inference Microservices (NIMs)

Job description

NVIDIA is seeking a Senior Applied Research Scientist to advance deep learning data curation and document extraction pipelines for foundation-model training. You will tackle multi-modal data (documents, images, audio, video) at petabyte scale, building efficient models, and contributing to the Nemotron Parse ecosystem.

You will collaborate with ML engineers and researchers to deploy NIMs, publish findings, and mentor juniors while serving as a key contributor to production-ready pipelines and