Technical Program Manager (AI Inference)

Location: San Francisco, CA; New York City, NY; or Seattle, WA. Hybrid policy requires staff to be in one of the offices at least 25% of the time.

Annual salary: $290,000–$435,000 USD

Company

Anthropic develops reliable, interpretable, and steerable AI systems intended to be safe and beneficial for users and society.

What you will do

• Own programs that scale Claude’s serving fleet across accelerator and cloud platforms from direction through delivery.
• Bring new capacity online across compute, infrastructure, security, networking, and inference teams.
• Manage trade-offs among capacity, utilization, cost, latency, reliability, and production demand.
• Lead reliability programs, large migrations, incident follow-ups, and inference readiness for model launches.
• Improve developer productivity, toolchains, development environments, CI/CD, and release processes across accelerator platforms.
• Coordinate roadmaps across production, research, safeguards, infrastructure, and cloud and silicon partners.

Requirements

• 7+ years of experience as a TPM or equivalent engineering/SRE lead owning large, multi-team infrastructure programs in production distributed systems, ML infrastructure, or cloud platforms.
• Technical fluency in serving infrastructure, request routing, load balancing, autoscaling, capacity planning, cluster orchestration, deployment, rollout, and observability.
• Working understanding of LLM inference hardware concepts, including prefill, decode, batching, caching, compute, memory bandwidth, and interconnects.
• Experience building program mechanisms from scratch and driving outcomes across organizations without direct management authority.
• Strong written communication and the ability to work with senior engineers, researchers, and cross-functional partners.
• Bachelor’s degree or equivalent education, training, or professional experience is required.

Nice to have

• Direct experience with LLM or ML inference serving at scale, including model servers, scheduling, KV caching, quantization, speculative decoding, or multi-model routing.
• Experience with GPU, TPU, or Trainium infrastructure and workloads spanning multiple accelerator platforms.
• Experience with multi-cloud or multi-region capacity management, demand forecasting, cost modeling, or compute unit economics.
• Experience with performance programs, ML compilers, kernels, numerics, collective communication, Kubernetes, or job scheduling.
• Practical use of AI tools, automations, or agents in program management and fleet operations.

Culture & Benefits

• Collaborative environment combining research, engineering, policy, and business expertise.
• Flexible working hours and office-based collaboration.
• Generous vacation and parental leave.
• Competitive compensation, benefits, and optional equity donation matching.
• Visa sponsorship is available, with reasonable efforts and immigration-lawyer support for eligible offers.