Skip to main content

Senior Fullstack · AI Engineer

Building

SaaS PlatformsAI AutomationsCustom Apps

Senior fullstack developer building SaaS platforms, AI automations, and custom apps. Remote from the Philippines.

Fullstack services

Automate business operations

Build agentic apps across marketing, sales, and support, designed to eliminate bottlenecks and grow with your business.

ReactNext.jsSupabasePostgreSQL

Senior Fullstack Developer

Custom app delivery with ReactJS, Next.js, FastAPI, PostgreSQL: architect, design, develop, deploy, secure, and scale.

  • React
  • Next.js
  • Supabase
  • PostgreSQL
  • Python
  • CRM
  • ERP
  • CMS
Pythonn8nZapierLangGraph

AI Application Engineer

Custom AI agents, RAG pipelines, LLM APIs, and workflow automations that connect your tools and reduce manual work.

  • Python
  • FastAPI
  • Agents
  • Chatbots
  • LLM
  • RAG
  • LangChain
  • LangGraph
  • n8n
WordPressWooCommercePHPMySQL

WordPress Development

Custom themes, plugins, and stores, with SEO and GEO built in.

  • Marketing Sites
  • Landing Pages
  • SEO
  • GEO
  • WooCommerce
  • Elementor

System architecture

SiteChat at a glance

Modular monolith with one main API: widget and merchant SPA talk to Bun + Hono; FastAPI owns crawl, embeddings, and pgvector search. Same shape I use for serious SaaS work.

Read the SiteChat case study →

How I ship

Spec-driven development with AI

I run SDD with Claude Code, Cursor, and Codex. Specs and architecture land first; agents implement against those contracts. Next to vibe coding in Replit, Lovable, or Base44, the difference shows up in what you can still maintain six months later, and then in cost.

  1. 1Write the spec and acceptance criteria
  2. 2Lock architecture boundaries
  3. 3Plan with Claude Code
  4. 4Implement with Cursor and Codex
  5. 5Ship via PR and deploy
  • You keep an owned codebase: reviewable PRs, tests, and deployable services, not a disposable prototype
  • Less rework: architecture and acceptance criteria are agreed before the bulk of the code exists
  • Lower token spend: agents stay on a written contract instead of rediscovering the system every chat