AI Maniacs

Building and maintaining AI Maniacs has been an ongoing exercise in curriculum design under change. Unlike a product spec or a library API, AI education has to be rewritten constantly — model names move, capabilities shift, and yesterday's best practice is today's dead end. That reality pushed me to design the content architecture for easy revision: small, single-topic pages with clear prerequisites, folder-based auto-navigation so new modules slot in without hand-edited menus, and a focus on patterns (agent design, MCP, workflow composition) that outlast specific tools. The Docusaurus + Docker + Caddy stack keeps the infrastructure boring so the work stays on content, not plumbing.

AI Maniacs shows up on my portfolio as proof that I can design and ship a non-trivial content product — not just an app, but a structured learning experience with a real audience and a maintenance burden that hasn't been abandoned. It also represents a shift in the kind of work I take on: from purely writing code for other teams to packaging expertise into reusable educational material. For anyone evaluating me for technical-content work, developer education, or roles that blend building and teaching, this is the clearest single example.