Manager, Software Engineering

NVIDIA•Durham, NC

4d•$224,000 - $431,250

About The Position

NVIDIA’s Managed AI Research Superclusters (MARS) team builds and scales the infrastructure, platforms, and tools used by researchers and engineers to develop AI/ML systems. The team operates the platform NVIDIA researchers use to train frontier models, including large GPU clusters spread across multiple clouds and partners. Its systems support pre-training, post-training, reinforcement learning, workload scheduling, data and checkpoint movement, and the underlying Kubernetes-based platform. The Manager, Software Engineering will lead an engineering team and own a core area of the AI infrastructure platform end to end, including its roadmap, architecture, reliability, and on-call health. The manager will grow a team of software and systems engineers, work directly with research teams to translate their needs into platform capabilities, and shape the future direction of the platform. The role also contributes to the shared Kubernetes-based platform codebase and its open-source path.

Requirements

BS or MS in Computer Science, Computer Engineering, or a related field, or equivalent experience.
12+ years of overall software engineering experience.
5+ years managing engineering teams, including leading tech leads or managers.
Track record of building and running large-scale distributed systems in production, such as clusters, platforms, or infrastructure services used by other engineers.
Hands-on depth in at least one of workload scheduling and resource management, distributed storage and data systems, or Kubernetes-based platforms.
Experience hiring and growing senior engineers and building teams that maintain high quality under pressure.
Experience owning reliability for critical services, including on-call, incident response, postmortems, and follow-up fixes.
Clear written communication and the ability to bring contested technical questions across several teams to a decision and document it.

Nice To Haves

Experience with GPU clusters for AI training at scale, including thousands of GPUs, large training runs, and high-speed interconnects.
Experience with reinforcement learning or post-training workloads and their infrastructure needs.
Contributions to open-source infrastructure projects or experience taking an internal platform open source.
Experience leading teams across time zones.
Uses AI coding and operations tools in daily work and has improved a team's productivity with them.

Responsibilities

Lead, hire, and grow a team of software and systems engineers, including senior and staff-level engineers and tech leads.
Own a core area of the AI infrastructure platform end to end, including its roadmap, architecture, reliability, and on-call health.
Work directly with model training teams to turn research needs into platform capabilities.
Establish clear ownership and decision rights, including named owners, written decisions, and clean interfaces with partner teams.
Drive engineering quality at fleet scale through automation, built-in observability, and safe rollouts across many clusters.
Partner with infrastructure, storage, networking, Kubernetes, and cloud-provider teams across NVIDIA, as well as external vendors and open-source communities.
Measure researcher-relevant factors such as reliability and latency and use them to set priorities.
Contribute to the shared Kubernetes-based platform codebase, including its open-source path, alongside the team's systems.