Eat'n -AI Chef

Inspiration

I picked Eitan Bernath's brief because our names sound almost identical — Eitan, Tan — and I genuinely love cooking. But like millions of people in Eitan's audience, I'm guilty of saving cooking videos I never actually make. The recipe is trapped inside the video — no ingredient list, no shopping list, no way to organize it. I wanted to build the app that bridges that gap. The name came naturally: swap the vowels in Eitan's name and you get Eat'n. It sounds like "eating" and "eating in" — cooking at home instead of ordering out.

What it does

Eat'n turns cooking videos into action. Paste any YouTube, TikTok, or Instagram cooking video URL and AI instantly extracts a full structured recipe — title, ingredients with exact amounts, step-by-step instructions, prep/cook time, servings, and difficulty. Every recipe auto-generates a checkable shopping list. Premium users can scan their fridge with the camera to detect ingredients and find matching recipes from their saved collection, ranked by match percentage. Recipes are saved into personal cookbooks that users can organize, favorite, and track. A digital cookbook marketplace offers curated recipe collections as one-time purchases.

How we built it

The frontend is React Native with Expo and TypeScript, using Expo Router for file-based navigation and React Context for state management. The backend is a FastAPI Python server hosted on Railway that handles video downloading (yt-dlp), audio transcription (OpenAI Whisper), and recipe extraction (OpenAI GPT-4 Vision). Authentication and the PostgreSQL database run on Supabase. All monetization — subscriptions and non-consumable purchases — is powered by RevenueCat's SDK. The app was built solo over the course of the hackathon, iterating from auth and navigation through to AI features and subscription flows.

Challenges we ran into

Getting RevenueCat product identifiers to match correctly between the dashboard and the SDK was tricky — the package type shortcuts (current.monthly) were selecting the wrong products, so we had to build a resilient fallback chain that matches by exact product identifier first. Keeping subscription state in sync between RevenueCat (source of truth for purchases) and Supabase (source of truth for in-app cancellations) required careful hydration logic on every app launch. Video analysis across three different platforms (YouTube, TikTok, Instagram) each had unique download and parsing challenges. And building a polished, feature-complete app as a solo developer under a tight deadline meant constant prioritization.

Accomplishments that we're proud of

The core flow works end-to-end: paste a video URL, get a full recipe with shopping list, save it to a cookbook — in under 30 seconds. The fridge scanning feature genuinely detects ingredients and matches them against saved recipes with percentage-based scoring. The monetization is real — three revenue streams (subscriptions, one-time cookbook purchases, and a future creator marketplace) all running through RevenueCat. The UI is clean and intuitive, designed specifically for the Gen Z/Millennial audience that lives on food content. And the name — Eat'n — ties the whole thing together.

What we learned

AI-powered content extraction is incredibly powerful but requires careful prompt engineering to get consistent structured output. RevenueCat's SDK is excellent for handling the complexity of in-app purchases, but syncing entitlement state with your own backend requires thoughtful architecture. Building a full-stack mobile app solo taught me to ruthlessly prioritize — ship the core value first, polish second. And most importantly: the gap between recipe inspiration and actually cooking is a real, unsolved problem that people care about.

What's next for Eat'n -AI Chef

Social features — share cookbooks with friends, follow other users, see what your friends are cooking. A creator platform where food influencers like Eitan can publish and sell their own digital cookbooks through the marketplace. Dietary preferences so users can filter recipes by vegan, gluten-free, keto during extraction. Meal planning with a weekly calendar and combined shopping lists. Smart grocery integration with services like Instacart to order missing ingredients directly. User analytics — tracking most-used ingredients, favorite creators, and cooking patterns to serve personalized recipe suggestions. And Android — the React Native codebase is cross-platform ready.

Built With

• expo.io
• fastapi
• openai
• python
• railway
• react-native
• revenuecat
• supabase
• testflight
• typescript