I build the paved roads other engineers ship through — self-service infrastructure, GitOps delivery, and cloud platforms that teams can actually maintain. My work centres on AWS: a multi-environment EKS delivery platform (TaskFlow) provisioned with Terraform and promoted across dev, staging, and prod through ArgoCD; an Internal Developer Portal on Backstage with a golden-path template that scaffolds a production-wired service in ninety seconds; a production-shaped observability platform on OpenTelemetry and the Grafana LGTM stack with trace-to-log correlation and SLO burn-rate alerting; and an incident simulation lab that puts real incident response and chaos engineering into practice. I document everything I learn publicly on GitHub, Hashnode, and LinkedIn — including the parts where I was wrong — because the best engineers make knowledge accessible.
Nigeria +234 806 444 4651
Designed and delivered TaskFlow, a production-grade, multi-environment Kubernetes platform on AWS EKS, and the Internal Developer Portal that sits on top of it. Provisioned all infrastructure with Terraform (VPC, EKS, ECR, RDS, OIDC, ALB, Route 53). Built GitHub Actions CI/CD pipelines with OIDC-based auth to AWS, ArgoCD GitOps for continuous delivery across dev/staging/prod, and a full observability stack with Prometheus and Grafana. Added a Backstage IDP with a golden-path template, TechDocs, and live Kubernetes and ArgoCD status per service — with the portal itself delivered by ArgoCD.
Built a containerized deployment platform on AWS ECS Fargate — deliberately scoped outside Kubernetes to demonstrate versatility. Implemented ECS deployment circuit breakers with automatic rollback, GitHub Actions CI/CD pipeline with pipeline health-check gates using aws ecs wait services-stable, and CloudWatch alerting. Documented the gap between API-accepted deployments and actual task health — a real-world reliability lesson.
An Internal Developer Portal on AWS EKS, delivered by ArgoCD. A software catalog spanning four repositories with ownership resolved through GitHub OAuth; TechDocs built in CI and served from S3 over IRSA — the external builder pattern, so the portal never runs a docs build itself; and live Kubernetes workload status plus ArgoCD sync state on every service page. The centrepiece is a golden-path Software Template: a developer fills a form and gets a production-wired repo with a Dockerfile, Helm chart, OIDC CI, and ArgoCD delivery in ninety seconds. Backstage itself runs under ArgoCD management from its own AppProject — delivered by the same GitOps pipeline it exposes.
Stack: Backstage · AWS EKS · Terraform · Helm · ArgoCD · GitHub Actions (OIDC) · ECR · RDS PostgreSQL · S3 · IRSA · Pod Identity
A production-grade GitOps delivery platform built for the TaskFlow application on AWS EKS. Provisioned all infrastructure with Terraform (VPC, EKS, ECR, OIDC, ALB, ExternalDNS, ArgoCD). Implemented a GitHub Actions CI/CD pipeline that promotes builds across dev, staging, and prod with manual approval gates, automated rollback on failed health checks, and per-stage Slack notifications. Enforced policy-as-code with OPA/conftest and secured deployments with ACM TLS, OIDC-based AWS auth, and Trivy container scanning.
Stack: AWS EKS · Terraform · Helm · ArgoCD · GitHub Actions · OPA/Conftest · Trivy · Prometheus · Grafana
A production-shaped observability platform for a containerized service on AWS EKS. Traces, logs, and metrics flow from an OpenTelemetry-instrumented Node.js app through an OTel Collector daemonset into the Grafana LGTM stack — Loki, Tempo, and Mimir — each backed by S3. Includes trace-to-log correlation (click a log line, jump straight to its trace) and multi-window SLO burn-rate alerting following the Google SRE pattern, routed through Alertmanager to Slack and email. S3 access is keyless via EKS Pod Identity — no static credentials. Provisioned with Terraform, installed with Helm.
Stack: AWS EKS · OpenTelemetry · Grafana · Loki · Tempo · Mimir · Alertmanager · Terraform · Helm · S3 · Pod Identity
A production incident simulation lab on AWS EKS. Eight real failure scenarios — from misconfigured deployments to a multi-fault cascading failure — each broken on purpose, diagnosed, fixed, and documented through a full Break → Detect → Fix → Improve cycle with real evidence: kubectl output, Prometheus queries, and Grafana dashboards. Built with Chaos Mesh for fault injection against a live, instrumented workload.
Stack: AWS EKS · Terraform · Chaos Mesh · Prometheus · Grafana · Helm
A real-world DevOps handover simulation on AWS ECS Fargate. Demonstrates container deployment, circuit breaker rollback, health-check pipeline gates, and CloudWatch alerting — without Kubernetes. Built to show that good DevOps practice applies across platforms, not just EKS.
Stack: AWS ECS Fargate · GitHub Actions · CloudWatch · ECR · ALB · Route 53 · Terraform