How I built a pipeline that takes a markdown blog post, processes mermaid diagrams into PNGs, deploys to my website, and automatically publishes to Substack. One git push, two platforms, zero manual steps.
How I replaced local development with a high-frequency cloud VM, a 2TB data disk, automated daily backups, and Terraform. My laptop is just a window into a machine that never sleeps.
Brad Feld built a 12-skill system to run his company through Claude Code. I manage multiple companies with layered CLAUDE.md files instead. Here is why simplicity wins when your system is already sophisticated.
Billing exports to BigQuery, budget alerts with smart deduplication, and Slack notifications - here's the four-step process I use to get full billing visibility across GCP projects and eliminate cost surprises.
Before DevOps pipelines, before CI/CD, before monitoring - I set up a data warehouse. Here's why centralizing data is the foundation for every decision a startup will make.
How I manage Slack channel memberships through YAML configuration. New employee joins? Add them to config, run a script, they're in all the right channels.
How I built a serverless alert system that uses Claude to analyze error logs and post human-readable summaries to Slack. No Docker, no servers, under $1/month.
How I connected Apple's App Store Connect webhooks to Slack, with automatic commit tracking between releases and forced auto-upgrades for internal TestFlight users. Build uploaded? Here's what changed, and your app will update itself.
For every company I work with, the first thing I build is a central knowledge base that becomes the DNA of the business. LLMs finally make this possible.
How I built a webhook-based integration that makes GitHub and Monday.com work as a single system. Developers commit code, and task statuses update automatically through the entire deployment pipeline.
The difference between vibe coding and structured AI-assisted development. How I use specs to guide AI through iterative, validated steps - not generate everything in one shot.
After years of deploying production infrastructure across AWS, Azure, and Google Cloud for 10+ startups, I'm launching OneOps.cloud to bring enterprise-grade DevOps to every early-stage company.
For the first time, machines can truly understand human language—not through brute force transcription, but because AI predicts what I'm going to say. This changes everything about how I interact with technology.