How we built a RabbitMQ deployment for EKS from scratch on the official Cluster Operator instead of a Bitnami chart, using a thin custom Helm chart and ArgoCD sync-waves.
How I designed a production observability platform using SigNoz on EKS with S3-tiered cold storage, KEDA-scaled OpenTelemetry collectors, zero-code Python APM via the OTel Operator, and encrypted RDS metadata — optimized for cost without sacrificing retention or trace depth.
How I deployed a self-hosted Stirling PDF instance on AWS EKS for a SaaS client — with OAuth2 SSO, persistent volumes, and ALB ingress sharing.
How I delivered a cost-optimized, auto-scaling GitHub Actions runner fleet on AWS using spot instances, pre-built AMIs, and time-based pool scheduling — eliminating idle compute costs for a SaaS client.
How I replaced VPN infrastructure with a Cloudflare Tunnel to provide both zero-trust VPC access via WARP and SSO-gated frontend application delivery through a private ALB — all on real public domains.
How I decomposed a monolithic Terraform state into isolated Terragrunt modules for a SaaS client — without touching live infrastructure.
How I designed an automated kill switch for a client's Athena data platform that disables OpenMetadata credentials within seconds of a scan threshold breach — no human intervention required.
A breakdown of how we deploy n8n in production on Amazon EKS using queue mode, Valkey for Redis-compatible brokering, and AWS Secrets Manager via External Secrets Operator.
GPU nodes on demand, zero cost at night, models that survive restarts — how I deployed self-hosted Ollama on Amazon EKS using a single Helm chart with Karpenter, KEDA, and ArgoCD.