DevOps and infrastructure work: deployment, automation, and the habits that keep systems running.
26 posts · all topics
Development environments lie in two directions. Clone production and it is too expensive to use. Mock it and it is confidently wrong. Here is a third option, and a working hybrid lab that joins a Kubernetes cluster on my laptop to a real Windows host behind a real AWS load balancer for about nine cents an hour.
Typing my automation handed me most of its rules for free; the type carries the contract. One rule came free with nothing, and it was the one my system was breaking: every file gets exactly one job that writes it. On the difference between rules you remember and rules the structure carries.
The best thing anyone told me about my own system this year was that I had not invented it. The scripts I grew by hand are a data pipeline, the shape is a DAG, and the discipline is solved. What the tools actually sell, why I am not buying yet, and the difference between wanting an engine and wanting a map.
A static site has no server to check a password, which is fine right up until the day you need one. My staging site is the problem in miniature. It is a full mirror of production, except it also carr...
A strict Content-Security-Policy on a static site allowlists inline scripts by hash. Change one script, forget to update the hash, and every page renders blank. The incident that taught me, and the deploy guard that ends it.
Not a tools list. The tools changed three times this year; what changed underneath is where the hours go: specs in the morning, batch review in the afternoon, and the judgment work that never left.
The code got cheap. The review did not. Here is the review workflow I actually use when AI agents write most of the diff: three passes, an adversarial second agent, and the discipline of reading what ships.
A practical, plain-spoken manual for building software with an AI agent: the setup, the loop, context management, and verified Claude Code and Copilot commands. Written like the old Linux HOWTOs, by someone who runs it daily.
DevOps was always a feedback loop. The first one ran between people. The one worth building now runs the system back on itself, so every failure and every change makes the next one cheaper and safer. The recursion is the compounding.
Microservices used to be a tax a solo builder could not afford. AI agents change that math. Here is why I am drawn to the architecture for what I build next, and the discipline that keeps it from becoming a mess.
This site was rebuilt from scratch by directing an AI agent, then migrated off cPanel onto AWS S3 and CloudFront. Here is what that actually looked like, the parts that worked, and the day a deploy script quietly deleted production.
A practical, repeatable workflow for building software with an AI agent: specify, delegate, verify, record. The loop I actually run every day, with the parts that keep it from turning into slop.
When a system gets too big to hold, almost everyone reaches for the same move: make it simpler. It is the wrong move. You cannot simplify the ocean. You can only chart it. This is about the difference between simplicity and legibility, why decisions quietly migrate to whoever holds the context, and how to make a complex system visible enough to navigate without pretending it is small.
I use GitHub Copilot at work and the Claude Code CLI at home, every day. Here is the firsthand comparison for DevOps and infrastructure work, from someone who ships with both, in the context each one actually lives in.
Life Ops is running your life as a deliberate system, designed the way you design software: observe before acting, choose responses over reactions, build systems that make the good path the easy one, and version the whole thing like code.
Agent mode is when you stop typing code and start directing an AI that reads your codebase, writes the change, runs the tests, and reports back. Here is what it actually is, how the loop works, and when to reach for it, from someone who ships production systems this way.
Antifragility is not resilience. Resilient things survive disorder; antifragile things gain from it. Here is how I apply Taleb's framework to fifteen years of engineering, the way I build films and books from code, and a ten-day journal cycle run as a stress rhythm.
DevOps is not a job title and never was. It is a thesis about how to build systems that can change without breaking. Fifteen years in, here is what actually compounded, what automation never solved, and what agent mode is now revealing about the practice.
AI hasn't replaced software engineers. It has moved the bottleneck. The cost of writing code dropped to near zero; the cost of deciding what to build, how to architect it, and when to ship became the whole job. A field report from a year of running production work in agent mode.
A 3:40 pop-EDM battle-rap. Plan 9 the Bell Labs bunny puts on his headphones at 03:14 UTC, slips into an AWS Availability Zone, and raps every service he passes (Lambda, S3, KMS, kubectl, EBS, Glacier) over a deadmau5-flavoured rework of Bach's Toccata in D minor. Three voice personas (lead Plan 9, autotuned female ad-libs locked to A4, deadpan whisper) plus a Transformer-droid layer in the gaps that delivers the Werner Vogels easter egg, "everything fails, all the time," and the universal devops punchline, "it's always DNS." Five passes from baseline to ship.
Layer upon layer, bricks to a skyscraper, deploy after deploy. What was done last year and what this year has become, and working furiously and peaceful at the same time.
On learning that creative eruption happens not as single explosion but as multiple parallel convection cells, discovering four domains where power deploys, and traveling 172 million kilometers while the red supergiant teaches conscious direction of energy
The DevOps Handbook is the clearest account I have read of why some engineering teams ship fast without breaking everything while others stay stuck in deployment hell. If you've ever wondered how companies like Netflix and Amazon deploy thousands of times per day without catastrophic failures, this book reveals the playbook.
There's a moment when you realize you're no longer programming computers. You're programming intelligence itself. Not through code, but through clear expression of intent, and that single shift changes everything about how you architect software, delegate complexity, and build systems you couldn't build alone.
This commit message tells a story that would have been impossible five years ago. A record of human-AI collaboration that produced something neither could have created alone. Git was built for human-paced development, but AI changes all of these assumptions.
Building enterprise AWS infrastructure in weeks with AI. CDK stacks, production systems, and the cognitive cost of compressed timelines.