Own horizontal functional slices across multiple layers of the infrastructure stack from hardware and kernel/runtime up through cluster scheduling, capacity control planes, and customer workloads. Grasp system-level trade-offs (such as compute utilization vs. memory headroom and reliability), surface hidden technical collisions early, and deliver risk assessments. Operate in an environment where static multi-year roadmaps have given way to rapid priority shifts and urgent escalations.
Establish disciplined intake gates, work-in-progress limits, and decision lock-in so programs don't stall. Replace repetitive status-reading meetings and manual toil with a single live source of truth, AI-assisted workflows, and concise decision artifacts that busy engineers and leads can review and approve rapidly. Replace reactive escalations with structured cross-organization partner intake, rigorous OKR/QBR cadences, and measurable infrastructure outcomes across compute efficiency, capacity yield, hardware launch velocity, and production reliability.
Job description provided by Google.
View the original posting
