Building in Public: Why I'm Starting This Blog
After years of building side projects in private, I'm starting a blog to document the process. Here's why, and what to expect.
Part 1 of the series: Building The Hub
For the past few years, I’ve been building an AI multi-agent system called The Hub. What started as a simple automation project has grown into a network of specialized agents — each with their own role, personality, and set of capabilities.
Why now?
I’ve been documenting my work privately in an Obsidian vault (also called The Hub), but I realized that some of the technical challenges I’m solving might be useful to others. Things like:
- How to coordinate multiple AI agents across different platforms
- Running MCP servers on a homelab for local-first AI tooling
- Building practical workflows with n8n, Claude, and Firebase
What to expect
This blog will be a mix of technical deep dives and higher-level reflections on building with AI. I’ll cover the architecture of The Hub, share lessons from my homelab setup, and occasionally write about frontend development and developer tools.
If you’re interested in practical AI engineering — not just theory, but real systems running in production — this is for you.
The stack
This blog itself is built with Astro, styled with Tailwind CSS, and hosted on Firebase. The design is inspired by blog.google — clean, card-based, and content-focused.
More posts coming soon. Subscribe to the RSS feed to follow along.