The Tools

Six tools. One workflow.

🧠

Claude

AI Thinking Partner & Builder

The core of everything. Not just code generation — architectural thinking, content writing, design decisions, debugging, and strategy. Claude is the first conversation before any line of code is written.

How I actually use it

I treat Claude as a co-founder who has no ego. I describe problems, not solutions. The output quality depends entirely on how well I ask the question.

Next.js

Frontend Framework

React-based framework for building multi-page websites. One file = one page. Shared components written once, used everywhere. Vercel (the company behind Next.js) makes deployment seamless.

How I actually use it

This website is built entirely in Next.js. Previously used plain HTML — the upgrade was necessary once we went beyond a single landing page.

🗄️

Supabase

Backend & Database

Postgres database + authentication + real-time + storage — all in one. The backend layer I never have to think about. Write SQL once, the API is auto-generated.

How I actually use it

Meeting request forms, dynamic data, anything that needs to persist. The RLS (Row Level Security) system means the database is safe to connect directly from the frontend.

🐙

GitHub

Version Control & Source

Every file, every change, every version — tracked. The source of truth for all code. Also the trigger that starts the deployment pipeline automatically.

How I actually use it

I push changes to GitHub → Vercel detects the push → site updates in 30 seconds. No manual deployment needed.

🚀

Vercel

Deployment & Hosting

The deployment platform made by the same team as Next.js. Connects to GitHub, builds and deploys automatically on every push. Free tier handles everything at this scale.

How I actually use it

Zero-config deployment. Connect once, forget forever. Every push to GitHub = live site update in under a minute.

💜

Lovable.dev

Rapid Prototyping

AI-powered frontend builder. Describe what you want in plain English, get working React code. Best for fast prototypes and exploring UI ideas before committing to production code.

How I actually use it

Early-stage experiments and UI exploration. When I want to see something quickly without writing code first.

The Workflow

How it all connects.

🧠
Claude
Think & design
✍️
Write code
Files created
🐙
GitHub
Push & store
🚀
Vercel
Auto-deploy
🌐
Live site
In ~30 seconds
Supabase sits separately — connected via URL + publishable key. Any page that needs to store or retrieve data talks to it directly. No server. No backend code to maintain. Just a database URL in the config.

Honest Lessons

What actually surprised me.

Lesson 1

Prompting is a skill, not a shortcut

The quality of AI output is entirely a function of how well you describe the problem. "Make a button" and "Make a primary call-to-action button in sage green that scales on hover with a subtle shadow" produce wildly different results. Investment in learning to prompt = compound returns.

Lesson 2

Components before pages

Build the navigation and footer first. These are shared. Everything else is unique per page. I wasted time on landing pages with copy-pasted nav because I didn't know about components. Never again.

Lesson 3

Your database schema is your architecture

Supabase tables are simple — but the decisions you make about structure shape everything downstream. Think about what data you're collecting and how you'll query it before creating the table.

Lesson 4

Ship → observe → fix → ship

The cycle time matters more than the perfect build. A live site with three bugs teaches you more than a perfect local version nobody sees. GitHub → Vercel makes this loop fast enough to be useful.

Lesson 5

Design system first, pages second

The most valuable thing I built was a consistent set of CSS variables (colours, fonts, radius, shadow) before touching any page. Changing the sage green now means changing one line, not 40.

Lesson 6

Domain expertise is the actual differentiator

An AI tool doesn't know what a real estate sales team needs. I do. The intersection of "knows the domain deeply" + "can now build using AI" is where the competitive moat lives.

What I've Built

Experiments from the playground.

Personal Website v1Single HTML landing page. Form connected to Supabase.
Shipped
Personal Website v2Full multi-page Next.js site. This site.
Live
Meeting Request FormSupabase table → RLS policy → live form. Responses in database.
Live
Substack RSS MirrorAuto-fetching Substack posts and displaying them in custom design via RSS2JSON.
Live

Curious about the process?

I write about building here — in the context of a non-coder founder navigating the AI engineering world.