Blog Post

In October 2025 I left Mozilla feeling behind on development skills. Eight months later I’m the ArgoCD SME at Jasper AI. Here’s what actually closed the gap.

What my commit history actually says about redesigning and shipping this site — Hugo, a Go backend on Cloud Run, and the messy middle.

19 ArgoCD-managed applications across 8 nodes: how my K3s cluster, GitOps workflow, network policies, and monitoring fit together.

How I monitor every layer of my homelab — bare metal, VyOS, Longhorn, Kubernetes, and a public VPS — from a single Grafana instance.

A free Kubernetes playground with a real cluster and a browser terminal — five tasks from ‘what’s a Pod?’ to deploying an app on the internet.

My Kubernetes learning platform silently broke for 24 hours — AI helped build it, harden it, and break it. What that taught me about staying in the loop.

Security fixes, monitoring additions, and a secret-management migration off KSOPS after the K8s Learn platform went live.

K8s Learn is a hands-on Kubernetes training platform I built and self-host: K3s, ArgoCD, n8n, Ollama, and Cloudflare Tunnel.

A visual tour of my homelab platform: Proxmox, K3s, GitOps with ArgoCD, Terraform, and services running across Docker and Kubernetes.

A mid-project update on standardizing Datadog instrumentation, monitors, and alerting at Jasper AI — and the SRE observability model behind it.

I’ve joined Jasper AI as a DevOps engineer, focusing on the foundational infrastructure that supports their continued scale.

How I used AI assistance to accelerate building a production-grade homelab — GitOps, Kubernetes, ArgoCD, and monitoring — while actually learning the fundamentals.