Resevia — AI Booking Agent for Beauty Salons
Resevia is an AI-powered booking agent that handles reservations and customer engagement for beauty salons — starting with salons, with broader service businesses to follow. This repo contains the public-facing marketing website and waitlist funnel, built to validate the product and capture early interest.
Live site: resevia.com
What This Is
A conversion-focused marketing site designed to explain the product, build trust, and capture leads via a waitlist. Key sections include:
- Hero — value proposition and primary CTA
- Problem / Solution — framing the pain restaurants face
- How It Works — step-by-step product walkthrough
- Industries — target verticals (beauty salons, service businesses)
- Social Proof & Reviews — trust-building
- Pricing Teaser — tiered plans preview
- Waitlist — lead capture with email confirmation via Resend
- Legal — Privacy Policy and Terms of Service pages
Tech Stack
| Layer | Technology |
|---|
| Framework | Next.js 14 (App Router) |
| Language | TypeScript |
| Styling | Tailwind CSS |
| Animations | Framer Motion |
| Database | Supabase (Postgres) — waitlist storage |
| Email | Resend — transactional email |
| Analytics | Microsoft Clarity, Meta Pixel, custom UTM/event tracking |
| Deployment | Vercel |
| Fonts | Montserrat (headings) + Inter (body) |
How It Was Built — Claude Code as Co-Pilot
This project was built primarily through natural language prompting with Claude Code — Anthropic's AI coding agent — rather than writing code by hand.
The approach:
- Describe, don't write — pages, components, and features were specified in plain English. Claude Code generated the implementation, handled wiring (API routes, Supabase schema, email flows), and iterated on design.
- Prompt-driven iteration — UI refinements, branding changes (logo, fonts, animations), and feature additions were all driven by follow-up prompts rather than manual edits.
- Agent autonomy — Claude Code managed file structure, resolved TypeScript errors, installed dependencies, and maintained consistency across components without manual intervention.
- Git history as a prompt log — each commit roughly maps to a prompting session or design direction change, showing how the site evolved from a blank canvas to a polished marketing funnel in a short time.
This workflow demonstrates how a non-engineering founder or solo developer can ship a production-quality web product at speed using AI-assisted development — writing minimal raw code while maintaining full ownership of architecture decisions and product direction.
Project Structure
/app # Next.js App Router pages (routes, API endpoints)
/api/waitlist # Waitlist form submission → Supabase + Resend
/pricing # Pricing page
/how-it-works # Feature walkthrough
/waitlist # Waitlist landing page
/privacy-policy # Legal
/terms # Legal
/components # Reusable UI components
/layout # Navbar, Footer
/sections # Homepage sections (Hero, HowItWorks, etc.)
/lib # Utility functions
/public # Static assets (logo, images)
Environment Variables
Sensitive credentials are kept out of the repo via .env.local (gitignored). To run locally, create .env.local with:
NEXT_PUBLIC_SUPABASE_URL=
NEXT_PUBLIC_SUPABASE_ANON_KEY=
RESEND_API_KEY=
What This Demonstrates
- Full-stack Next.js app with server-side API routes
- Supabase integration for persistent data storage
- Transactional email with Resend
- Analytics instrumentation (UTM tracking, custom events, Clarity, Meta Pixel)
- Animated SVG branding and polished UI with Framer Motion
- Production deployment on Vercel
- AI-assisted development workflow using Claude Code
