Senior Manager, EC2 Nitro, EC2 Nitro

EC2 Nitro drives the planet"s largest, fastest growing, and most feature-rich compute cloud, and EC2 Machine Learning Systems (MLS) is where that platform meets the most demanding AI workloads on Earth. We build and scale the compute platforms behind AWS"s largest ML training and inference clusters - integrating custom hardware, firmware, systems software, and services to turn AI accelerators (NVIDIA GB200/GB300/VeraRubin-class UltraServer/UltraCluster) into reliable, high-performance capacity for customers ranging from startups to the largest AI labs in the world.

We are looking for a seasoned engineering leader to own an existing, at-scale ML compute platform and to set and drive its multi-year technical strategy. As a Senior Manager leading a team of 3-5 managers and roughly 40-50 engineers, you will be responsible for the health, velocity, and long-term direction of an organization that spans hardware bring-up, firmware, cluster management, and application software. You will define where the platform needs to be several years out, make the architectural and investment bets that get us there, and build the org - leaders, senior engineers, and mechanisms - capable of delivering it.

In this role, you will operate at the intersection of a broad and deep set of technical teams. You will partner at a senior level with silicon and hardware partners (Annapurna and NVIDIA) to co-design and bring up NVIDIA GB200/GB300 platforms and UltraServer/UltraCluster systems, and with science, ML framework, and other AWS service organizations to remove ambiguity, align roadmaps, and land capabilities that no single team could deliver alone. You will be measured not only on what your organization ships, but on the clarity of the technical strategy you set and your ability to influence direction well beyond your immediate teams.

The ideal candidate:

• Has a deep technical background in systems software - sufficient to earn the trust of senior engineers and make sound architectural and design judgments in a Linux environment involving C, C++, and/or Rust. You will not be writing production code day-to-day, but you must be able to go deep, ask the right questions, and hold a high technical bar.
• Brings strong operational instincts for large-scale software delivery - version control, automated build and release, and software quality practices - and knows how to instill them across an organization.
• Has in-depth knowledge of ML frameworks and cluster management, and understands the workloads and failure modes of large-scale ML training and inference at cluster scale.
• Has a track record of setting multi-year technical strategy, driving org-level execution through other managers, and influencing partners across organizational boundaries.