The opportunity
As a Technical Program Manager on the Compute team, you will help drive the planning, coordination, and execution of programs that keep Anthropic's compute infrastructure running efficiently at scale. Our compute fleet is the foundation on which every model training run, evaluation, and inference workload depends.
What you'll do
Own and drive critical programs across the compute lifecycle, coordinating: execution across multiple engineering, research, and operations teams
Build and maintain operational visibility into the compute fleet, ensuring: the organization has a clear picture of supply, demand, utilization, and health
Lead cross-functional coordination for compute transitions: bringing new capacity online, migrating workloads, and managing decommissions across cloud providers and hardware platforms
Partner with engineering and research leadership to navigate competing: priorities and drive alignment on how compute resources are planned, allocated, and used
Identify and close operational gaps across the compute pipeline, whether: through new tooling, improved processes, or better cross-team communication
Own trade-off discussions between utilization, cost, latency, and: reliability, synthesizing inputs from technical and business stakeholders and communicating decisions to leadership
What they're looking for
- Develop and improve the processes and frameworks the team uses to plan,: track, and execute compute programs at increasing scale and complexity
- Have 7+ years of technical program management experience in infrastructure,: platform engineering, or compute-intensive environments
- Have led complex, cross-functional programs involving multiple engineering: teams with competing priorities and ambiguous requirements
- Have experience working with research or ML teams and translating their needs: into operational plans and technical requirements
