Senior Applied Research Scientist

NVIDIA•Santa Clara, CA

About The Position

The Data and Agent Loop Lab is seeking a Senior Applied Research Scientist to build multimodal search and curation systems for Physical AI. The role focuses on developing systems that help researchers and customers find, filter, and surface high-quality data across video, lidar, radar, text, and other modalities. Areas of work include embedding-based retrieval, quality-scoring models, active-learning loops, and relevance-evaluation frameworks. The scientist will develop datasets, benchmarks, and experimental frameworks to guide data curation for robotics and autonomous-vehicle applications, and publish findings for the broader community. The role also translates research into production-grade software and produces papers, blog posts, and documentation to help customers apply the methods. The work is conducted with researchers and engineers on large-scale data challenges and supports customers building Physical AI pipelines on NVIDIA's platform.

Requirements

M.S. or Ph.D. in a relevant field such as machine learning, information retrieval, robotics, autonomous vehicles, or multimodal AI, or equivalent experience.
At least 8 years of industry or research experience.
Deep understanding of data challenges specific to Physical AI.
Strong software-engineering fundamentals and experience shipping production-grade systems, not only research prototypes.
Clear written and verbal communication skills, including the ability to write papers and explain tradeoffs to non-expert collaborators.

Nice To Haves

Experience mentoring junior researchers or engineers.
Hands-on experience with data curation or search in robotics, autonomous vehicles, or embodied AI.
Familiarity with model-training pipelines and the role of data quality in model performance.
Experience with large-scale video data collection, processing, and retrieval.
Prior work on semantic or neural search, including dense retrieval, CLIP-style embeddings, or multimodal indexing.
Publications in leading venues on data-centric AI, information retrieval, or multimodal learning.

Responsibilities

Research and design multimodal search and curation systems for Physical AI data, including video, lidar, radar, and text.
Develop retrieval, data-quality scoring, active-learning, and relevance-evaluation approaches to help users find and surface high-quality data.
Develop datasets, benchmarks, and experimental frameworks that establish data-curation guidelines for robotics and autonomous-vehicle contexts.
Publish research findings to benefit the community.
Translate research into production-grade software.
Write papers, blog posts, and documentation that help customers understand and apply the team's methods.
Stay current with advances in data curation, information retrieval, and multimodal learning, and bring relevant insights to the team.