Cukr

Never Forget a Meal Again.

Comment

Inspiration

We noticed that people constantly lose track of recipes they find on social media, forget what they planned to cook, and end up staring into their fridge wondering what to make. The kitchen should be a place of joy, not stress. We wanted to build a companion that keeps your cooking life organized — from discovery to the dinner table — so you never forget a meal again.

What it does

Cukr is an AI-powered cooking companion built around four core tabs:

Explore — A daily cooking dashboard that shows what you're cooking today, upcoming meals for the week, and personalized "For You" recipe suggestions tailored to your taste.

AI — Four powerful features: (1) type your ingredients and get instant recipe suggestions, (2) paste any link from YouTube, TikTok, or Instagram and extract the full recipe with an AI-generated image, (3) receive 5 personalized recipe recommendations every day, and (4) describe any recipe in your own words and have AI format it beautifully.

Calendar — A weekly meal planner where you schedule recipes, see your agenda at a glance, and get automatic shopping lists from your plan.

Search & Collections — Instantly search recipes and organize them into custom collections (Favorites, Italian Night, Family Recipes, etc.) so nothing is ever lost.

On top of this, Cooking Mode provides a step-by-step swiping experience. When you tap a recipe, Cukr checks your ingredients — if you have everything, you jump straight into hands-free cooking. If you're missing items, you can plan it for later, which adds missing ingredients to your shopping list and sends you a reminder notification the day before so you have time to prep and shop.

How we built it

Cukr is built with Expo (React Native) for a cross-platform mobile experience, starting iOS-first. On the backend, we use Firebase Cloud Functions to handle API logic, recipe processing, and AI orchestrations. For authentication, we chose Clerk for its seamless sign-up/login flow and user management. Monetization is handled through RevenueCat, which manages subscriptions and in-app purchases across platforms.

For AI-powered recipe extraction from social media links (YouTube, TikTok, Instagram), we integrated Apify to scrape and parse content from those platforms, then pass it through our AI pipeline to structure the recipe data. We use RecipeDB to source accurate ingredient images so users always see real, recognizable visuals when checking their pantry or browsing recipes.

Notifications are handled with Expo's local notification system, which powers the day-before reminders that alert users to prep and shop for missing ingredients. The landing page itself is built with React, TypeScript, Tailwind CSS, and shadcn/ui components.

Challenges we ran into

Extracting recipes reliably from social media was tricky — each platform structures content differently, and video-based recipes (TikTok, YouTube) don't have a standard recipe format. Getting Apify scrapers to work consistently across platforms and then having AI accurately parse unstructured content into clean ingredient lists and steps took significant iteration.

Managing the ingredient-checking flow — where the app compares what a user has against what a recipe needs, then intelligently routes them to either Cooking Mode or Plan for Later — required careful state management across multiple screens in Expo.

Scheduling local notifications with Expo that reliably fire the day before a planned meal, across different timezones and device states, also presented edge cases we had to work through.

Accomplishments that we're proud of

The end-to-end cooking flow works seamlessly: find a recipe, check ingredients, cook step-by-step or plan for later with automatic reminders. It feels like a real cooking companion, not just a recipe database. URL extraction from social media actually works — paste a TikTok or YouTube link and get a fully structured recipe with an AI-generated image. This solves a real pain point for millions of people who save recipes on social media and never find them again. RecipeDB integration gives us accurate, real ingredient images that make the ingredient-checking experience feel polished and trustworthy. The subscription model via RevenueCat is clean and ready for monetization from day one.

What we learned

Clerk made authentication significantly easier than rolling our own — we saved weeks of development time on auth flows. Apify is powerful for web scraping but requires careful error handling for each platform's quirks. We learned to build fallback parsing logic for when scrapers return incomplete data. Local notifications in Expo are reliable for simple scheduling, but we'll likely need to move to push notifications via Firebase Cloud Messaging as we scale to handle more complex reminder patterns. Using RecipeDB for ingredient images taught us that data quality matters more than data quantity — having accurate, recognizable images for common ingredients makes the whole experience feel more trustworthy.

What's next for Cukr

Launch the Android version (currently iOS-first) Migrate to Firebase Cloud Messaging for push notifications to support more advanced reminder patterns Add voice input for hands-free recipe creation while cooking Build social features — share collections and meal plans with family and friends Integrate with grocery delivery services for one-tap ingredient ordering Expand multi-language AI support for a global audience Add smart seasonal suggestions based on local ingredient availability Explore Apple Watch companion for cooking timers and quick recipe glances

Share this project:

Updates