Rechef - Turn saved recipe into cooked meal

• 

• 

Inspiration

I grew up in a kitchen. My mom is a chef, and her recipe book was the center of our home. Every meal came from those worn, marked-up pages. She taught me how to cook, and I genuinely loved it.

But life changed. As a busy student, cooking slipped away. Takeouts became the default. I'd scroll Instagram and see incredible recipe videos, my mom would even send me some, and I'd save them with the best of intentions. But those saved videos would vanish into a graveyard of random bookmarks, buried under memes, reels, and everything else. And with ADHD, the moment I'd open my saved folder to find a recipe, three other videos would pull my attention away before I ever made it to the kitchen.

When I read Eitan's brief, "help people actually cook what they've been meaning to make," it hit home. That's exactly the problem I live with every day. The inspiration is everywhere. The follow-through is what's missing.

What It Does

Rechef turns saved recipes into cooked meals.

It bridges the gap between "I saw this" and "it's on the table" by removing every barrier between saving a recipe and actually cooking it.

Save from anywhere. Import recipes from YouTube, TikTok, Instagram, websites, or even a photo of a handwritten recipe. Rechef uses AI to extract the full recipe, ingredients, steps, and all, so you never have to scrub through a 10-minute video again just to find the measurements.

Organize effortlessly. Recipes are saved into Cookbooks, customizable collections that keep everything structured. Built-in smart cookbooks like Shared With Me and Pantry Picks (which recommends recipes based on what you already have at home) mean you always know what to cook next.

Know what you need. Every recipe shows which ingredients are already in your Pantry and which are missing. Add missing items to your Grocery List with a tap, or buy everything on Instacart with one click, without leaving the app.

Schedule and get reminded. Plan when you want to cook and Rechef sends you a notification when it's time, so recipes don't get forgotten again.

Cook with confidence. When you start cooking, Rechef enters Cooking Mode, a distraction-free, step-by-step guided experience with built-in timers and a real-time AI Cooking Assistant. Stuck on a step? Not sure what "fold in" means? Just ask. It's like calling your mom for help, except she's always available.

Share and discover. Share any recipe as a link. Friends can view it, save it, and cook it themselves. Recipe creators can share their collection through Rechef and see engagement analytics, opening the door for influencers to distribute recipes beyond a video description box.

Track nutrition. Every recipe includes AI-generated nutritional facts for the health-conscious cook.

How We Built It

Research First

Before writing a single line of code, we talked to people, a lot of them. Recipe content creators, busy moms, college students, working professionals, and fitness enthusiasts. The pattern was strikingly consistent: people love finding recipes, but the journey from saved to cooked is full of friction. The most common pain points were:

• Recipes scattered across multiple platforms with no central place to organize them
• No easy way to figure out what ingredients they already have vs. what to buy
• Forgetting about saved recipes entirely
• Feeling lost mid-cook when video instructions move too fast or assume expertise

These insights directly shaped our feature set and prioritization.

Technical Architecture

Mobile App: Flutter (Dart) We chose Flutter for its ability to deliver a beautiful, native-feeling UI on both iOS and Android from a single codebase. The app follows Clean Architecture with a feature-based folder structure and uses Riverpod for state management and GoRouter for navigation. Key integrations include:

• RevenueCat SDK for subscription management (Rechef Pro: monthly/yearly plans)
• Firebase Auth with Google Sign-In and Sign in with Apple
• Share Extension so users can import recipes directly from any app's share sheet
• Speech-to-text for hands-free interaction with the AI assistant while cooking
• Local notifications for cooking reminders and timers

Backend: Node.js/TypeScript with Express Our API handles the heavy lifting of recipe extraction and AI processing. It runs on Express 5 with PostgreSQL (via Drizzle ORM) and uses a background job queue for async processing. Core capabilities include:

• AI-powered recipe extraction using OpenAI GPT-5.2 for parsing videos, images, and web pages into structured recipe data
• Video frame analysis with FFmpeg to extract food-relevant frames from recipe videos before sending them to the AI for processing
• Streaming AI chat via Server-Sent Events for the real-time cooking assistant
• Content parsing for YouTube (transcript extraction), TikTok, Instagram, and general websites (Cheerio + schema.org recipe extraction)

Web App: Next.js A lightweight Next.js app powers shared recipe links, allowing anyone to view a recipe in the browser, even without the app installed.

Instacart Partnership We integrated with the Instacart Connect API so users can send their grocery list directly to Instacart and check out with one click. This also creates a referral-based revenue stream. We earn a percentage on groceries purchased through the use of our app.

Monetization with RevenueCat

Rechef uses a freemium model powered by RevenueCat:

• Free tier: 5 recipe imports per month, full access to cooking mode, pantry, and grocery lists
• Rechef Pro: Unlimited imports, AI cooking assistant, advanced nutrition facts, and priority processing

RevenueCat handles subscription entitlements (Rechef Pro), paywall presentation, and customer management. It's initialized on app launch and linked to each user's Firebase UID.

Additional revenue stream: The Instacart referral integration provides a commission on grocery purchases made through the app, a recurring, usage-based revenue channel that scales with engagement.

Challenges We Ran Into

• Balancing the competition with real life. Building an MVP in four weeks alongside work and school meant every hour counted. Prioritization was everything.
• Research and validation. Reaching people in the cooking niche, conducting interviews, and distilling dozens of pain points into a focused feature set took more time than expected, but was absolutely worth it.
• The Instacart partnership. Integrating with a third-party commerce API involved back-and-forth coordination, API key provisioning, and edge case handling for different product catalogs.
• AI reliability. Extracting structured recipes from wildly inconsistent video and web formats required extensive prompt engineering and fallback strategies. A TikTok recipe is very different from a blog post with schema.org markup.
• Polish under pressure. Making the cooking mode feel seamless, with timers, step navigation, and an AI assistant all working together, required tight iteration in a short window.

Accomplishments We're Proud Of

The accomplishment we're most proud of is simple: Rechef works, and we use it every day.

Since building the app, I've been cooking more, spending less on takeout, and eating healthier. That's not a pitch line, it's my actual life now. The app I built to solve my own problem is solving it.

Beyond the personal win, we're proud of:

• A fully functional AI pipeline that can extract a recipe from a TikTok, YouTube video, Instagram post, website, or photo, and turn it into a structured, cookable format in seconds
• A seamless cooking experience with step-by-step mode, integrated timers, and a AI assistant
• A working Instacart integration that lets users go from recipe to grocery checkout
• A sharing system that gives recipe creators a clean, dedicated way to distribute their content

What We Learned

The biggest lesson: removing barriers changes behavior. When saving a recipe is more organized, when missing ingredients appear automatically, when a notification reminds you to cook, and when an AI walks you through every step, you actually cook. It sounds obvious in hindsight, but experiencing the behavior change firsthand was powerful.

We also learned that the recipe-to-meal pipeline is a surprisingly underserved space. Social media has made recipe discovery effortless, but everything after discovery, organizing, shopping, scheduling, and cooking, is still manual and fragmented. There's a real opportunity to own that entire journey.

On the technical side, building a production-quality Flutter app with a robust AI backend under time pressure sharpened our skills enormously. Working under a deadline with a clear goal is the best motivator there is.

What's Next for Rechef

Immediate (post-hackathon):

• Gather user feedback and squash remaining bugs
• Go live on the App Store and Google Play
• Expand AI accuracy across more recipe formats and languages

Short-term growth:

• Partner with recipe content creators to import their catalogs and share via Rechef links
• Launch social media channels featuring recipes powered by Rechef
• Introduce a creator incentive program, rewarding influencers based on engagement with their shared recipes (views, saves, cooks)

Long-term vision:

• Meal planning with smart suggestions based on pantry, dietary preferences, and budget
• Social features: cook-alongs with friends, family recipe books, community cookbooks
• Expand Instacart integration to additional grocery delivery partners globally
• Become the platform where recipes live after they're discovered, the place where inspiration becomes dinner

Built With

• cheerio
• drizzle-orm
• express-5
• extension
• ffmpeg
• firebase-auth
• flutter-(dart)
• flutter-local-notifications
• google-sign-in
• gorouter
• gpt-4o-mini)
• instacart-connect-api
• next.js
• node.js
• openai-api-(gpt-5.2
• postgresql
• react
• revenuecat-sdk
• riverpod
• server-sent-events-(sse)
• share
• sign-in-with-apple
• speech-to-text
• tailwind-css
• typescript
• winston