Senior Kubernetes Platform Engineer for AI Compute

anthropic; Greater London; Hybrid; GBP 325.000 - 485.000; Vollzeit

Zusammenfassung

Anthropic is seeking a highly senior systems engineer to own and extend the Kubernetes scheduler for our accelerator fleets and to scale the control plane beyond current limits.

You will design and operate core cluster services, build controllers and CRDs, and partner with research to translate workload needs into platform capabilities, while collaborating with cloud providers on features and on-call incident response.

Qualifikationen

• Significant software engineering experience building and operating production distributed systems.
• Proficiency in at least one systems language (Go, Python, Rust or C++).
• Deep, hands-on Kubernetes experience beyond user level (scheduler, controllers, apiserver, multi-tenant clusters).
• Ability to debug complex issues across stack from API to node network root causes.
• Track record of designing for reliability and clear failure semantics in critical systems.
• Strong written and verbal communication, ability to build consensus with stakeholders.

Aufgaben

• Own, operate, and extend the Kubernetes scheduler for Anthropic's accelerator fleets.
• Scale the Kubernetes control plane to support clusters beyond typical limits.
• Design, build, and operate core cluster services like service discovery.
• Build and maintain custom controllers, operators, and CRDs.
• Collaborate with research, training, and inference to translate workload requirements into platform capabilities.
• Work with cloud providers on required features and escalations.
• Participate in on-call and drive incident response and postmortems to avoid repeats.

Kenntnisse

Go

Python

Rust

C++

Kubernetes

Debugging

Communication

System design

Jobbeschreibung

Anthropic is seeking a highly senior systems engineer to own and extend the Kubernetes scheduler for our accelerator fleets and to scale the control plane beyond current limits.

You will design and operate core cluster services, build controllers and CRDs, and partner with research to translate workload needs into platform capabilities, while collaborating with cloud providers on features and on-call incident response.