Case Study
Family Planner
A build-vs-buy experiment: could AI tools let me ship a full-stack app faster than evaluating existing solutions? AI-powered recipe import from TikTok, YouTube, and blogs, drag-and-drop meal planning, and smart grocery lists — built end-to-end with Next.js, Supabase, and Claude API.
Tech Stack
Problem
My family needed a better way to save recipes from TikTok, YouTube, and blogs — and plan meals from them. But this was also an exploration: with AI-native tools, is it faster to build exactly what you need than to evaluate and compromise with existing apps?
Approach
Treated this as a real build-vs-buy decision. Scoped the user need (recipe capture, meal planning, grocery lists), then built end-to-end using AI-assisted development — Next.js and Supabase for the platform, Claude API for intelligent recipe extraction from any URL or photo. The goal was to test how fast a product leader with AI tools can go from idea to shipped product.
Outcome
Shipped a fully functional family platform in a fraction of the time traditional development would require. Open source — and a proof point that AI-native building changes the calculus on build-vs-buy decisions for product leaders.
Key Highlights
- AI recipe import from TikTok, YouTube, Instagram, blogs, and photos
- Drag-and-drop meal planner with smart grocery lists
- Full-stack with auth, database, and real-time sync