Boriss N

Designed and built a GPU Kubernetes-as-a-Service platform supporting 100+ H100/H200 nodes across bare metal and multi-cloud, used by 20+ teams for AI training and LLM inference. Achieved $410K in annual savings through cost optimization and migration of OpenShift clusters to on-prem.

Implemented FluxCD-driven Container-as-a-Service on AKS/EKS with Kyverno and Workload Identity for governance. Developed Python tooling to analyze cost data from Kubecost deployments and automated Azure provisioning with Terraform, cutting deployment timelines by 50%.

Led Anthos Service Mesh (Istio) implementations across financial services GKE clusters, establishing reference patterns for clients. Created a GKE design template adopted as corporate standard and introduced Argo Rollouts for zero-downtime deployments.

Managed a global hybrid OpenShift platform with 100+ nodes across on-prem and AWS, including cross-region networking and secure data paths for ML workloads. Migrated SDN from OpenShift SDN to Cilium for eBPF observability and tighter network policies.

Responsible for data centre operations and server administration, including IBM, HPE, and Dell EMC hardware, and virtualisation with VMware and Hyper-V.

Performed system administration duties including data centre operations, server administration, and virtualisation.

Grade: 2:1

Earned a Bachelor of Engineering in Computer Systems & Networks with a 2:1 classification.

Grade: A

Completed an HND in Internetworking Technologies with a Grade A.