SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Salary: USD 290,000 - 435,000 / annual
Anthropic is seeking a Technical Program Manager to drive programs that scale Claude's inference serving infrastructure. The role spans four closely coupled areas: scaling and infrastructure (routing, autoscaling, capacity management for the serving fleet), velocity (toolchains, development environments, CI/CD for inference engineers across accelerator platforms), inference engine (software between accelerator kernels and routing layer), and performance (improving throughput, latency, cost, and reliability with correctness as a release criterion).
This is not a status-tracking role. You will own outcomes end-to-end, often where no single team is accountable for the whole picture. Your value comes from crossing boundaries into Research, Safeguards, Security, Compute, Infrastructure, API, and cloud and silicon partners, bringing back requirements and dependencies that no team sees from inside its own space. You will build the minimum machinery the work needs and remove it when it stops helping. You will have Claude as a collaborator and are expected to push hard on what AI-assisted program management and fleet operations can look like.
Key responsibilities include: owning programs that scale Claude's serving fleet across accelerator and cloud platforms from early direction through delivery; driving new capacity online end-to-end across compute, infrastructure, security, networking, and inference teams; building visibility into supply, demand, and utilization with capacity and finance partners; leading reliability programs and large migrations without disrupting production or research; driving programs that make inference engineers faster by defining productivity metrics, finding bottlenecks, and improving how software is built, tested, and released; coordinating roadmaps and priorities across production, research, and safeguards needs; tracking performance and efficiency work from opportunity to measured impact in production; and owning inference readiness for model launches.
Anthropic's mission is to create reliable, interpretable, and steerable AI systems. The Inference organization serves Claude to millions of users and enterprise customers across Anthropic's products and the world's largest cloud platforms. The stack is deep and tightly coupled: accelerator kernels, an in-house inference engine managing the entire token path for production and research workloads, distributed routing, autoscaling, and capacity management. The company builds across GPUs, TPUs, and Trainium, and complexity grows with every platform added.
Minimum qualifications: 7+ years as a TPM or equivalent engineering/SRE lead role, owning large multi-team infrastructure programs end-to-end in production distributed systems, ML infrastructure, or cloud platforms; technical fluency in large-scale serving infrastructure including request routing, load balancing, autoscaling, capacity planning, cluster orchestration, deployment, rollout, and observability; working understanding of how LLM inference uses hardware (prefill and decode, batching, caching, compute vs. memory bandwidth vs. interconnect) or track record of getting deep in unfamiliar technical domains quickly; track record of building program mechanisms from scratch on ambiguous, fast-moving efforts and retiring process that has stopped earning its keep; ability to drive outcomes across organizations you don't manage with competing priorities and unclear decision ownership; strong written communication for plans, reviews, and executive updates; low ego and high ownership; care about Anthropic's mission and serving frontier AI safely and reliably.
Preferred qualifications: direct experience with LLM or ML inference serving at scale (model servers, batching, scheduling, KV caching, quantization, speculative decoding, multi-model routing); experience with accelerator infrastructure (GPU, TPU, Trainium) including cluster bring-up, hardware qualification, or running workloads across multiple accelerator platforms; experience managing capacity across multiple clouds or regions including demand forecasting, headroom policy, cost modeling, or unit economics for compute-heavy services; experience with developer productivity, release engineering, or CI/CD at scale for hardware-dependent workloads; experience running performance or efficiency programs with grounding in rooflines, kernels, ML compilers (XLA, Triton, Neuron), numerics, or collective communication; familiarity with Kubernetes-based orchestration and job scheduling; experience working with research teams and turning their needs into infrastructure requirements; practical experience using AI tools in your own workflow including building automations or agents.
