Revnut

The Problem

There's no "Famous Birthdays" for cars. If you want to look up a specific generation of a car — say, the E46 BMW M3 or the NA Miata — you're bouncing between Wikipedia, forums, and YouTube. There's no single place where enthusiasts can browse, compare, and rank every model ever made.

What It Does

Revnut is a community-driven car database with social voting:

• Browse by brand, model, generation — structured data for every car, organized the way enthusiasts actually think about them
• Community rankings — users vote on cars across categories (best looking, best sounding, best driver's car), and the rankings reflect the community's taste
• Rich editorial content — brand stories, model histories, and generation deep-dives written in an enthusiast voice

The site is built for discovery. You come for the E30 M3, you stay because you found a Lancia Delta Integrale you didn't know existed.

How It's Built

Next.js 16 (App Router) with Supabase for the database and Cloudflare Workers for hosting. The site generates 80K+ pages from structured data — every brand, model, and generation gets its own page with SEO-optimized content.

Images are served via Cloudflare Images. The voting system uses Supabase real-time subscriptions. Editorial content is stored as MDX and rendered server-side.

What I Learned

Revnut taught me SEO at a scale I'd never touched before. Building 80K+ pages that actually rank means thinking about structured data, internal linking, and content quality at a systems level — not just page by page.

It also taught me that the car enthusiast community is incredibly passionate and opinionated. The ranking system works because people care deeply about these cars. That emotional investment is the product's moat.