Senior Multimodal Data Curation Scientist | Remote

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

Job summary

NVIDIA’s Curator team seeks a Senior Applied Research Scientist to research, develop, and deploy deep learning models at scale for data curation and document extraction. You will work on multimodal data across documents, images, audio, and video, building scalable pipelines for foundation-model training.

The role emphasizes state-of-the-art data curation, reproducible experiments, and collaboration with ML engineers to scale pipelines in production using NIMs.

Qualifications

• Master's, Ph.D. or equivalent in data curation, document AI, information retrieval or multimodal research.
• Hands-on experience with computer vision and document-extraction models including OCR, layout analysis, table/figure extraction.
• 10+ years of experience developing multimodal systems across models and platforms; IR experience is a plus.

Responsibilities

• Develop efficient and performant models and data pipelines for multi-modal data extraction used in foundation model training.
• Build petabyte-scale extraction and content deduplication pipelines (document/html parsing, deduplication, semantic/substring deduplication).
• Scale pipelines to production via NVIDIA Inference Microservices (NIMs) and blueprints for deployment.
• Write papers, blog posts, documentation, and training materials for customers.
• Stay current with data curation advances in academia and industry.

Skills

Python programming

PyTorch

Multimodal data

Data curation

Distributed data frameworks

Machine learning pipelines

Education

Master's degree or PhD or equivalent in data curation/document AI/information retrieval/multimodal research

Tools

Ray

Spark

Dask

NIMs (NVIDIA Inference Microservices)

Job description

NVIDIA’s Curator team seeks a Senior Applied Research Scientist to research, develop, and deploy deep learning models at scale for data curation and document extraction. You will work on multimodal data across documents, images, audio, and video, building scalable pipelines for foundation-model training.

The role emphasizes state-of-the-art data curation, reproducible experiments, and collaboration with ML engineers to scale pipelines in production using NIMs.