7 days ago
Responsibilities
- Own, operate, and extend the Kubernetes scheduler for accelerator fleets, including scheduling plugins and policies for gang scheduling, topology awareness, and preemption.
- Scale Kubernetes apiserver, etcd, and controller-manager for extremely large clusters and identify future bottlenecks.
- Design, build, and operate core cluster services such as service discovery.
- Build and maintain custom controllers, operators, and CRDs.
- Partner with research, training, and inference teams to translate workload requirements into platform capabilities.
- Collaborate with cloud providers on required features and escalations.
- Participate in on-call, lead incident response, and develop postmortems, runbooks, and SLOs.
Requirements
- Significant experience building and operating production distributed systems.
- Proficiency in at least one systems language such as Go, Python, Rust, or C++.
- Deep hands-on Kubernetes experience involving the scheduler, controllers, apiserver, or large multi-tenant clusters.
- Ability to debug complex issues from API behavior through node- and network-level root causes.
- Experience designing for reliability, correctness, and clear failure semantics.
- Strong written and verbal communication and consensus-building skills.
- Preferred: Kubernetes internals or contribution experience, cluster schedulers or batch systems, control-plane or coordination-system scaling, ML infrastructure, GCP or AWS, GKE or EKS internals, infrastructure as code, Linux kernel tuning, cgroups, eBPF, and 12+ years of relevant industry experience including leadership of large ambiguous infrastructure projects.
Benefits
- Annual compensation range is £325,000–£485,000 GBP.
- Hybrid policy currently expects staff to work from an office at least 25% of the time, with some roles requiring more.
- Visa sponsorship is available for eligible roles and candidates.
- Benefits include competitive benefits, optional equity donation matching, generous vacation and parental leave, flexible working hours, and office collaboration space.
Tech Stack
Categories
About Anthropic
We're an AI research company that builds reliable, interpretable, and steerable AI systems. Our first product is Claude, an AI assistant for tasks at any scale. Our research interests span multiple areas including natural language, human feedback, scaling laws, reinforcement learning, code generation, and interpretability.