Written up from the episodes
Blog
The episodes, written down: real commands, real output and the reasoning behind each fix.
2026-10-27 · 9 min read
AI Engineer Roadmap 2026: What to Learn and What to Skip
A 7-stage AI engineer roadmap from a working platform engineer. Foundations, LLMs, RAG, evals, agents, production and security, each built and run for real.
2026-10-27 · 8 min readKarpenter on EKS: Pending Pods to a Right-Sized Node in 30s
A real Karpenter 1.15 run on Amazon EKS. Five pending pods, a NodePool with a hard CPU limit, one right-sized node in about 30 seconds, then consolidation.
2026-10-27 · 8 min readCrashLoopBackOff and OOMKilled (Exit 137) in a Real Incident
We broke a checkout service on a real Kubernetes cluster. CrashLoopBackOff, OOMKilled exit 137, a rollback that didn't fix it, and the service selector trap.
2026-10-27 · 9 min readSRE Interview Questions 2026 with Strong Answers
The SRE interview questions that keep coming up in 2026, from SLOs and error budgets to exit code 137, burn rate alerts and postmortems, with answers that land.