2 days ago
Base Salary
$224k - $357k/yr
Responsibilities
- Own the 3D data engine by sourcing, curating, filtering, and balancing large-scale image and video corpora for spatial reasoning.
- Build annotation and auto-labeling pipelines for 3D-grounded supervision, spatial questions, reference frames, correspondence, camera motion, and geometric attributes.
- Manage end-to-end data quality through deduplication, automated scoring, coverage analysis, and sampling strategies.
- Run and validate full pre-training and supervised fine-tuning pipelines, ensuring reproducibility and diagnosing data, checkpoint, throughput, and loss regressions.
- Build and operate 3D and spatial evaluation suites using public benchmarks, NVIDIA VANTAGE-Bench, and internally designed benchmarks.
- Partner with research scientists to translate 3D hypotheses into dataset experiments, ablations, and model improvements.
- Operate multi-node GPU clusters and tune data throughput, sharding, and dataloader performance.
- Ship results into Cosmos releases and appropriate open-source datasets and benchmarks.
Requirements
- MS or PhD in Computer Science, Electrical or Computer Engineering, Robotics, or a related field, or equivalent experience.
- 12+ years of experience building deep learning systems in Python with PyTorch or JAX on Linux.
- Deep expertise in 3D computer vision, multi-view geometry, structure-from-motion or SLAM, depth and camera pose estimation, point clouds, or 3D reconstruction.
- Hands-on experience building training data and evaluations for vision-language models and improving visual grounding and reasoning quality.
- Experience building large-scale multimodal data pipelines for distributed video and image processing, deduplication, captioning, annotation, automated quality metrics, and dataset versioning.
- Experience running and validating large-model training on multi-GPU and multi-node clusters, including distributed training and sharding strategies such as data, tensor, and pipeline parallelism or FSDP.
- Experience designing rigorous benchmarks, clean ablations, and evaluations resistant to gaming.
- Strong written and verbal communication and experience partnering with research scientists.
- Preferred qualifications include a PhD or publications at CVPR, ICCV, ECCV, NeurIPS, ICLR, or CoRL; web-scale or petabyte-scale video corpus experience; distributed infrastructure experience; 3D foundation model and auto-labeling experience; vision-language post-training, supervised fine-tuning, chain-of-thought data, reward modeling, or reinforcement learning experience; and evaluation infrastructure experience with tools such as VLMEvalKit.
Benefits
- Competitive base salary with equity and benefits.
- Applications are accepted at least until November 3, 2026.
- The position is an existing vacancy.
Tech Stack
Categories
About Nvidia
Nvidia designs and sells GPUs and accelerated computing platforms for data centers, AI/ML, graphics, gaming, and automotive, monetizing through hardware, software platforms (CUDA, AI frameworks), and systems like DGX and networking. Customers include cloud providers, enterprises, researchers, and OEMs. Founded in 1993 and headquartered in Santa Clara, it is a public company traded on NASDAQ under NVDA.