Staff Software Engineer, Kubernetes Platform at Anthropic

London, UK

on-site · full-time · Visa sponsorship

Apply for this role at Anthropic

Responsibilities:
- Own, operate, and extend the Kubernetes scheduler for accelerator fleets, including custom scheduling plugins, gang scheduling, topology-aware placement, and preemption policies.
- Scale the control plane components (apiserver, etcd, controllers) to support object and node counts far beyond typical limits; proactively identify and eliminate bottlenecks.
- Design, build, and operate core cluster services (service discovery, networking primitives) that every workload depends on.
- Develop and maintain custom controllers, operators, and CRDs to expose platform capabilities to consumers.
- Partner with research, training, and inference teams to translate workload requirements into scheduling and platform features.
- Collaborate with cloud providers on required features and escalations.
- Participate in on-call rotations, lead incident response, and drive reliability processes (SLOs, runbooks, postmortems).

Requirements:
- Significant experience building and operating production distributed systems at scale.
- Proficiency in at least one systems language (Go, Python, Rust, or C++).
- Deep, hands-on Kubernetes experience (scheduler, controllers, apiserver, or operating large multi-tenant clusters) beyond typical user-level knowledge.
- Proven ability to debug complex issues across layers (API, control plane, node, and network).
- Demonstrated design for reliability, correctness, and clear failure semantics in shared systems.
- Strong written and verbal communication; ability to build consensus across stakeholders.
- Preferred: experience with scheduler internals, controller-runtime, and scaling large control planes.