Senior On-Premise Platform Engineer

- Design and build on-premise infrastructure hosted in leased UK data centres - Architect and operate bare-metal Kubernetes clusters (control plane + workers) - Own compute, networking, storage, Linux OS, and platform architecture - Design platforms capable of 99.99% availability - Plan and execute capacity management, failover, and disaster recovery - Operate GPU-enabled infrastructure for AI inference and training - Build systems suitable for NHS 999 and emergency communications workloads - Design and maintain GitLab CI/CD pipelines (build, test, deploy) - Automate infrastructure provisioning (Terraform/IaC) - Automate Kubernetes deployments - Automate AI model and application releases - Implement GitOps workflows - Own day-2 operations, including upgrades, patching, and rollbacks - Minimise deployment risk in safety-critical environments - Build and operate Kafka-based streaming platforms - Support sub-second latency event processing - Design for traffic spikes, back-pressure, and failure scenarios - Ensure predictable behaviour under 999 call surges - Optimise systems for latency, throughput, and resilience - Operate production AI inference platforms (KServe, Seldon, Triton, or similar) - Enable GPU scheduling, isolation, and concurrency controls - Support model versioning, retraining pipelines, and lifecycle management - Implement canary releases - Implement versioned deployments - Implement safe rollback paths - Build observability using Prometheus, Grafana, and centralised logging - Define and monitor SLIs, SLOs, latency, uptime, and error budgets - Lead incident response and root-cause analysis - Implement least-privilege access, secrets management, and audit controls - Harden platforms for NHS, telecoms, and regulated environments