Staff Software Engineer, AI Runtime

about 5 hours ago

Mountain View, CA, USA or San Francisco, CA, USAStaff+

H1B Sponsor

Base Salary

$190k - $265k/yr

Responsibilities

Drive the architecture and evolution of AIR's managed GPU training platform.
Solve complex problems in large-scale training, including multi-node orchestration and GPU scheduling.
Enhance GPU efficiency and training performance across diverse model architectures.
Build resilience and observability foundations for multi-node jobs.
Collaborate with product, research, and platform teams to improve developer experience.
Lead end-to-end engineering efforts from design to production rollout.
Make high-impact contributions to core systems and support new accelerators.
Mentor engineers and influence Databricks' technical direction in AI training.

10+ years of experience in large-scale distributed systems, particularly in GPU training infrastructure.
Hands-on experience with distributed training frameworks like PyTorch and DeepSpeed.
Strong understanding of training resilience patterns for long-running jobs.
Solid grasp of GPU performance fundamentals and bottlenecks affecting training throughput.
Experience with managed, multi-tenant platform products in the cloud.
Strong foundation in algorithms, data structures, and system design.
Proven ability to deliver complex, high-impact initiatives.
Excellent communication skills for collaboration across teams.
Strategic mindset with a passion for mentoring engineers.
BS in Computer Science or a related field; MS or PhD preferred.