Staff Node Infra Engineer: Cloud Compute for AI

anthropic; Greater London; Hybride; GBP 325 000 - 485 000; Plein temps

Résumé du poste

Anthropic is seeking a senior Infrastructure Engineer to own the node lifecycle infrastructure for its AI compute fleet. You will design the strategy for ingesting, bringing up, health checking, and automated repair of GPU/TPU/Trainium nodes across clouds, enabling reliable experiments and scalable model training.

You will partner with research, product, and cloud teams to build robust orchestration, improve fault tolerance, and drive operational excellence in incident response and on-call

Qualifications

• Deep expertise in distributed systems and cloud platforms.
• Strong proficiency in at least one systems language (Go/Python) and IaC with Terraform.
• Hands-on experience with ML accelerators (GPUs/TPUs/Trainium).
• Track record leading cross-team technical initiatives spanning multiple systems.
• Ability to communicate with senior stakeholders.

Responsabilités

• Own the technical strategy and roadmap for node lifecycle management (ingestion, bring-up, health checks, automated repair).
• Drive cross-team initiatives to scale AI clusters across clouds and accelerator families.
• Design and operate systems that detect, isolate, and remediate unhealthy hardware automatically.
• Define infrastructure architecture to solve hard problems, directly or through others.
• Collaborate with cloud providers and internal teams to shape long-term compute and data strategy.
• Establish and evolve operational excellence practices (incident response, postmortems, on-call).
• Mentor and coach engineers to grow their technical skills.

Connaissances

Distributed systems

Kubernetes

Go

Python

Terraform

Cloud providers

Machine learning accelerators

Formation

Bachelor’s degree

Outils

AWS

GCP

Azure

Kubernetes internals

Mesos

Borg-like systems

Linux kernel

Description du poste

Anthropic is seeking a senior Infrastructure Engineer to own the node lifecycle infrastructure for its AI compute fleet. You will design the strategy for ingesting, bringing up, health checking, and automated repair of GPU/TPU/Trainium nodes across clouds, enabling reliable experiments and scalable model training.

You will partner with research, product, and cloud teams to build robust orchestration, improve fault tolerance, and drive operational excellence in incident response and on-call