The original problem
The first version of the site treated the blog as both a tutorial library and a personal writing feed. That made navigation messy. A user looking for help and a developer looking for engineering notes were pushed into the same list.
What changed
I split the site around roles instead of page types. Tools became the execution surface. Docs became the help center. Blog became a personal engineering zone. The URLs stayed under one domain, but the content boundaries became explicit.
Why static pages still worked
The site did not need a CMS to prove the structure. A first-pass static version was enough to settle the IA, internal linking, metadata, and SEO boundaries. That made it possible to move fast without blocking on a larger platform decision.
What I would improve next
- Introduce shared header and footer templates to reduce repeated static markup.
- Add structured content data so hubs can be generated instead of manually curated.
- Decide which legacy tutorial URLs should stay indexed and which should redirect cleanly to docs.