03 — Skills
What the pager taught me
Platform and SRE engineer with about eight years of building and running cloud-native systems end to end: Kubernetes platforms, GitOps delivery, high-throughput Kafka pipelines and the backend services on top.
Certifications
CKA · 2020CKS · 2021Azure Solutions Architect · 2021Azure Network Engineer · 2022GitHub Advanced Security · 2023
Kubernetes & platform
GKEAKSEKSRancherOpenShiftRKE2HelmArgoCDMulti-tenancyAir-gappedGPU workloads
Cloud
GCPAzureAWSApigeeAlloyDBDatastreamLanding zones
Distributed systems & data
KafkaStrimziPostgreSQLElasticsearchMongoDBRedisMySQL
Observability & SRE
PrometheusGrafanaLokiMimirTempoELKSLO/SLIIncident responseRCAKubecost
CI/CD, IaC & security
TerraformGitHub ActionsAzure DevOpsJenkinsGitLab CISAST/SCANeuVectorSSO automation
Languages & backend
GoJavaSpring BootPythonNode.jsC (eBPF)ReactNext.jsGraphQLREST
Highlights
Cloud migrationA US ISP, re-platformed onto GCPLead architect for the move from AWS: dual 300–400-node GKE clusters for data and applications, with GitOps through ArgoCD for 1000+ developers.
APIsApigee Edge to X, with nobody logged outMigrated 225+ microservices with zero downtime and no forced re-logins by using token exchange to carry sessions across.
DataKafka at 5–8M records a secondProduction Kafka with Datastream and AlloyDB for change-data-capture and low-latency serving.
PlatformA multi-tenant Kubernetes platformBuilt on Rancher with full tenant isolation, NeuVector security, Grafana, Loki, Mimir and Tempo, Kubecost showback and automated SSO across every tool.
DeliveryPaved-road CI/CDShared pipeline templates with mandatory security gates: SAST, SCA, secrets and IaC scanning. Moved 500+ applications from GitLab to GitHub onto them.
ObservabilityLogs and analytics at TB scaleAn ELK platform over 9–12 TB that keeps dashboards under 3 seconds, a Kafka and Logstash pipeline with PII masking, and a custom cold-storage query engine to cut retention cost.
SecurityAir-gapped and GPU KubernetesSecure clusters with no internet access for sensitive workloads, and GPU-operator setups for AI workloads on Rancher.
BackendA case-management platform at scaleArchitecting a Spring Boot and GraphQL (Netflix DGS) backend with a React workflow-authoring UI, built for about 20M active cases.
TeamA cloud-native team of 15+Built and led the team. Every engineer earned CKA or CKAD within six months, through real project work.