Staff Cloud Native Engineer — AI GPU Infra Architect

Nscale; Greater London; On-site; GBP 110,000 - 170,000; Full time

Job summary

Nscale is hiring a Staff Cloud Native Software Engineer to build, operate, and improve the cloud-native software integrations that connect AI applications and networking components at scale.

In this role you’ll work on shared Kubernetes-based platforms, deployment patterns, observability foundations, infrastructure architecture, and operational tooling that help internal teams run services safely and efficiently on GPU-backed infrastructure.

Qualifications

• At least 8 years of experience in production-level software development.
• Deep hands-on experience building and operating Kubernetes-native software: custom controllers, operators, CRDs, or admission webhooks — using controller-runtime, client-go, or equivalent.
• Strong understanding of Kubernetes internals: the API server, informer/lister patterns, reconciliation loops, and the object model.
• Strong networking fundamentals — CNI, service mesh, kube-proxy/eBPF datapaths, DNS, load balancing — and experience building software that integrates with these systems.
• Proficiency in Go (strongly preferred) or a similar language, with a track record of shipping well-tested, production-quality code at scale.
• Experience with observability practices — metrics, tracing, structured logging — built into software rather than added afterward.
• Comfortable owning components independently end-to-end, from design through operation, while setting direction for adjacent teams.
• Experience with or strong interest in GPU-backed infrastructure and AI workload patterns is a plus.
• Track record of leading technical design at a staff level and mentoring engineers through practical technical guidance.

Responsibilities

• Design, build, and operate Kubernetes-native software — controllers, operators, CRDs, and admission webhooks — that connects AI applications with core networking components on GPU-backed infrastructure.
• Extend Kubernetes control-plane capabilities to support AI workload requirements, including network policy controllers, CNI/service-mesh integrations, and resource/scheduling extensions.
• Own significant components end-to-end and set the technical direction for how they're designed, deployed, and operated across the team.
• Build reconciliation loops, informers, and client-go–based tooling that keep infrastructure state consistent between the API server, networking systems, and AI runtime components.
• Develop operational tooling and automation that make Kubernetes-native services easier for internal teams to deploy, run, and support.
• Drive infrastructure architecture decisions around how AI applications and networking components integrate across the platform, weighing trade-offs at a cross-team level.
• Build observability foundations for controller and operator software — metrics, structured events, tracing, and status reporting surfaced through the Kubernetes API and platform dashboards.
• Design systems that degrade gracefully and self-heal, using controller patterns (reconciliation, backoff, status conditions) to reduce manual intervention.
• Debug and resolve complex issues spanning the Kubernetes control plane, networking (CNI, service mesh, kube-proxy/eBPF datapaths), and workload runtime behavior on GPU-backed infrastructure.
• Define standards for safe rollout of controller and platform changes, including versioning, compatibility, and staged deployment.
• Set technical direction for how the team builds Kubernetes-native software, establishing patterns for controller design, CRD schema evolution, and testing strategy.
• Lead design discussions and code reviews, holding a high bar for Kubernetes API conventions and idiomatic client-go usage.
• Partner with platform engineering, infrastructure, and product teams to translate real developer and operational needs into clean CRDs, APIs, and controller-managed abstractions.
• Define reusable patterns, shared libraries, and scaffolding that let other teams build correctly on the platform without reinventing integration logic.
• Mentor engineers in Kubernetes internals, controller-runtime patterns, and sound operational judgement.

Skills

Kubernetes expertise

Go programming

Software leadership

Code reviews

Education

Bachelor's degree in Computer Science or equivalent

Tools

controller-runtime

client-go

CRDs

admission webhooks

Job description

Nscale is hiring a Staff Cloud Native Software Engineer to build, operate, and improve the cloud-native software integrations that connect AI applications and networking components at scale.

In this role you’ll work on shared Kubernetes-based platforms, deployment patterns, observability foundations, infrastructure architecture, and operational tooling that help internal teams run services safely and efficiently on GPU-backed infrastructure.