📰 Medium · LLM

The winners are not the flashiest copilots. They are the tools that reduce deployment risk, incident noise, Kubernetes cost, security… Continue reading on Mediu

A few years ago, I started the 100 Days of DevOps series, and it helped thousands of engineers strengthen their DevOps skills. Continue reading on Medium »

audit is a three-layer code analysis package that runs entirely inside Claude — no scripts to execute, no CI plugins to configure, no… Continue reading on Mediu

Our CI eval gate failed anything under 90%. A change landed at 91%, passed, and shipped a regression that cost us two days. The absolute… Continue reading on Me

Announcing mcp-xray: one graded report for any MCP server’s token cost, tool confusion, and surface bloat. Continue reading on Medium »

Multi-tenant SaaS agents have to tell every downstream API which tenant the current call belongs to. Distributed systems rely on… Continue reading on Medium »

We moved our source of truth from Framer to Git so content could be reviewed, versioned, and published like code. Continue reading on NEXT AI Engineering »

Devcontainers are great for reproducible, isolated dev environments. The one thing they tend to ruin? The polished, app-like coding… Continue reading on CodeX »

I have a two-node Harvester cluster and It runs VMs, manages storage and handles whatever workloads I throw at it. Continue reading on Medium »

Most enterprises already run MCP servers — one for the data warehouse, one for the ticketing system, one for the internal search index… Continue reading on Medi

A practical hands-on guide to building, testing, containerizing, and deploying an enterprise-ready MCP server with Python, FastMCP, Docker… Continue reading on

A practical, narrative-driven guide to Terraform, Infrastructure as Code, providers, state, modules, drift, CI/CD, security, OpenTofu, and… Continue reading on

A 10-part series on building software delivery systems that actually work. Continue reading on Medium »

This week dives into the future of IaC in the AI era, the unglamorous craft of module design, hard-won lessons on isolating Terraform… Continue reading on Mediu

There was a moment, sometime after midnight, when a 24-year-old software engineer named Arjun abandoned our app. Continue reading on Medium »

A 10-part series on building software delivery systems that actually work. Today: the six core pieces every harness is built from. Continue reading on Medium »

How platform teams are turning chaotic deployments into a calm, repeatable engineering discipline. Continue reading on Medium »

What we actually changed in the workflow, what broke, and why verification became the real bottleneck. Continue reading on Medium »

A guide to schedule Nomad jobs to run overnight AI work and return desktops before the workday starts with Ansible. Continue reading on HashiCorp Solutions Engi

If your team uses Jira for issue tracking and GitLab for CI/CD, you are sitting on a goldmine of automation potential. Continue reading on Medium »

Ray is an open-source distributed computing framework that makes it easy to scale Python workloads from a single machine to a full cluster. Continue reading on

A practical guide to wiring OpenClaw’s built-in OTel diagnostics into New Relic — so you can see exactly which agents and models are… Continue reading on Medium