No "proficient in Microsoft Office" filler. These are the tools, languages, and disciplines I use in active projects — listed honestly.
This site is its own portfolio piece. No CMS, no framework, no build step — here's the actual stack.
Hand-written PHP templates with shared includes for head, header, footer, and schema. All content — projects, services, skills — lives in JSON data files. Updating the site means editing a JSON file and uploading it. No database, no admin panel to hack.
A single hand-written stylesheet (~26KB) — no Tailwind, no Bootstrap. About 5KB of vanilla JavaScript handles the nav, the typewriter animation, scroll reveals, lazy video loading, and the contact form. Nothing else ships to the browser.
Every page outputs a JSON-LD schema graph (Person, WebSite, Breadcrumb), canonical URLs, and OG/Twitter meta. The site carries a maintained sitemap.xml, robots.txt, and llms.txt for AI search engines — the same GEO setup I build for clients.
Hosted on my own DirectAdmin VPS running OpenLiteSpeed on AlmaLinux — the same server stack I manage for client sites. SSL, security headers, fail2ban, caching rules, and 301 hygiene all handled in-house. Lighthouse scores live in the footer.
I also build in React when a project needs component-based UI, WordPress when a client needs a familiar CMS, and native mobile — Kotlin / Jetpack Compose in Android Studio and Swift in Xcode. The stack serves the project, not the other way around. Curious about the AI side? See my AI stack.
I favor minimal dependencies and infrastructure I can fully understand and explain. I don't use frameworks I can't read and debug at the source level. I don't build things that require specialized knowledge to maintain after I'm gone.
PHP + JSON for most sites. Not because it's trendy — because it's readable, portable, and doesn't require a build step to deploy. A client can open the files in any text editor and see exactly what's happening.
AI tools are part of my workflow, but they're tools. I use Claude primarily for structured output generation, schema drafting, and batch research tasks — not for writing code I don't understand or making decisions I haven't thought through.