Stack

Agents, code, and infrastructure.

Claude Code, Codex, Cursor. TypeScript. The hosting and WordPress layer I already run in production.

AI & Agents

  • Claude Code
  • Codex
  • Cursor
  • OpenAI
  • PythonRAG · evals
  • Gemini
  • OpenRoutermodel routing
  • MCP serverstool integrations

Code & Data

  • TypeScript / JavaScript
  • React / Next.js
  • HTML / CSSTailwind
  • PHP
  • SQL / MySQLInnoDB tuning
  • Supabaseauth · Postgres · pgvector
  • Git / REST APIs

Hosting & Infrastructure

  • Kinstamanaged WordPress
  • Cloudwaysmanaged cloud
  • DigitalOcean
  • AWSS3 · EC2
  • CloudflareDNS · R2 · CDN
  • Vercel
  • DNS / SSL / cachemigrations & debugging

WordPress & CMS

  • WordPresscore, multisite, debugging
  • WooCommerce
  • MemberMousemembership platforms
  • ACFcustom fields
  • Gravity Forms
  • Elementor
  • Custom pluginssmall plugins

Support Operations

  • Technical account management
  • SaaS implementation
  • Escalations / SLA
  • QA & bug reproduction
  • Docs & playbooks
  • Customer success

01—How I use AI

How I work with agents, and what I have already put in production.

Agentic coding

Claude Code, Codex, and Cursor run the loop: inspect, plan, implement, test, review. Each repo has instructions so the agent follows the same architecture and commands I would.

AI in production

I put custom AI chat on WordPress membership sites. It has to fit the site they already have.

AI systems engineering

Structured outputs, retrieval with citations, evals, traces, cost, model routing. The goal is systems that still work after launch.