Senior AI Data Platform Engineer

2100 NVIDIA USA; Santa Clara (CA); On-site; USD 152,000 - 288,000; Full time

Benefits offered by this job

Equity

Benefits

Job summary

NVIDIA is building the MARS data platform to support AI research at scale. The role focuses on designing and implementing an AI data layer across multi-cloud GPU fleets, with emphasis on efficient data placement, caching, and parallel I/O.

You will work with researchers to optimize training workflows and storage behavior across thousands of GPUs. You’ll contribute to Kubernetes-native storage APIs and ensure reliable, scalable data services with quotas and lifecycle management, impacting

Qualifications

• 5+ years building production systems or data-infrastructure software.
• Deep experience with distributed file systems or object storage at scale.
• Strong programming in Go, Rust, C or C++, plus Python for tooling and analysis.
• Proven ability to own a service in production and improve it.
• Clear written and spoken communication with engineers and researchers.

Responsibilities

• Design and build the AI data layer for a multi-cloud GPU fleet: datasets, checkpoints, and scratch data placement, caching, tiering, and movement between storage backends.
• Build high-throughput data paths for training and RL, including streaming loaders, distributed caches, and checkpoint save/restore at thousands of GPUs.
• Make data a first-class part of the compute platform.
• Create Kubernetes-native APIs, operators, and storage integrations for job data declarations, with deployment on Slurm and Kubernetes.
• Own quotas, capacity, and data lifecycle across hundreds of research teams with attribution and forecasting.
• Measure data usage and build I/O telemetry and analytics to guide placement and fixes.
• Own reliability: on-call, incident analysis, and changes to prevent data incidents.

Skills

Go

Rust

C/C++

Python

Distributed systems

Communication

Tools

Lustre

WEKA

VAST

GPFS

Ceph

S3

Job description

NVIDIA is building the MARS data platform to support AI research at scale. The role focuses on designing and implementing an AI data layer across multi-cloud GPU fleets, with emphasis on efficient data placement, caching, and parallel I/O.

You will work with researchers to optimize training workflows and storage behavior across thousands of GPUs. You’ll contribute to Kubernetes-native storage APIs and ensure reliable, scalable data services with quotas and lifecycle management, impacting