About the role

Anthropic's Inference organization is responsible for serving Claude to millions of users and enterprise customers with the speed, reliability, and efficiency that frontier AI demands. As we scale across multiple accelerator platforms—GPUs, TPUs, and Trainium—the complexity of our development environment grows in lockstep. We're looking for an Engineering Manager to build and lead a new team focused on developer productivity within Inference: making every engineer in the org dramatically more effective at building, testing, and shipping inference software.

This is a leadership role at the intersection of infrastructure and developer experience. You'll own the toolchains, workflows, and feedback loops that Inference engineers depend on every day. Your work will establish priorities to keep our engineering velocity high and driving investments to keep the larger org productive. You'll partner closely with Anthropic's central Infrastructure organization (where company-wide developer productivity efforts live) while ensuring the Inference org's unique needs are met.

This role is ideal for someone who gets deep satisfaction from unblocking other engineers, who thinks in terms of systems and feedback loops, and who can lead a team that operates at the seam between ML infrastructure and software engineering productivity.

Responsibilities:

• Build and lead a high-performing team focused on developer productivity for the Inference organization, hiring engineers who combine infrastructure expertise with a service-oriented mindset
• Own accelerator toolchain management across GPU (CUDA), TPU, and Trainium platforms—keeping compilers, drivers, libraries, and frameworks current, compatible, and well-tested so that Inference engineers can focus on model serving rather than environment issues
• Build infrastructure for efficient accelerator usage during development—including devbox environments, automation for pre- and post-landing validation, and shared tooling that reduces the friction of working across heterogeneous hardware
• Establish and drive productivity metrics across the Inference org, creating dashboards, alerts, and processes that surface slowdowns early (e.g., smoke tests red for extended periods, build times regressing, toolchain breakages) and ensure rapid resolution
• Identify and eliminate inefficiencies across Inference engineering workflows—proactively finding bottlenecks, toil, and friction points that slow down the org, and building systems or driving process changes to address them
• Partner with Anthropic's Infrastructure org to align on company-wide developer productivity initiatives, contribute Inference-specific requirements, and avoid duplicating effort while ensuring Inference's specialized needs (multi-accelerator support, large-scale testing) are well-served
• Coach and develop engineers on your team, providing clear direction, actionable feedback, and growth opportunities in a fast-moving environment

You may be a good fit if you:

• 3+ years of engineering management experience, ideally leading infrastructure, platform, or developer productivity teams
• Strong technical background in systems engineering, build/test infrastructure, or ML infrastructure—you can go deep on toolchain issues, CI/CD pipelines, and developer workflow optimization
• Experience managing toolchains or development environments for compute-intensive workloads (ML training/inference, HPC, large-scale distributed systems)
• Familiarity with at least one accelerator ecosystem (CUDA/GPU, TPU, or Trainium/AWS Neuron) and an appetite to learn the others
• A track record of defining and using engineering metrics to drive organizational improvement—you've built dashboards, set SLOs on developer workflows, or led initiatives to measurably improve engineering velocity
• Experience partnering across organizational boundaries—you know how to advocate for your team's needs while contributing to shared infrastructure efforts
• Strong communication skills and the ability to influence without authority across a technical organization
• A genuine passion for making other engineers more productive, and the empathy to understand their pain points deeply

Strong candidates may also:

• Experience with ML compiler toolchains (XLA, Triton, NeuronX) or accelerator driver/firmware management at scale
• Background in building or running shared development environments (devboxes, remote development, ephemeral environments) for hardware-dependent workflows
• Experience with CI/CD systems at scale, particularly for workloads involving accelerator hardware
• Familiarity with Kubernetes-based development and job scheduling environments
• Prior experience in a developer productivity or platform engineering role at a fast-growing AI/ML company