Inspiration
Standing in front of a refrigerator stocked with groceries while feeling like there is "nothing to cook" is a frustrating daily reality. Most cooking apps push rigid, hyper-restrictive diet plans or tedious calorie and macro tracking that turn mealtime into a chore. We built PortableCook to bring intuitive creativity back to cooking—using multi-modal AI to transform whatever ingredients you already have into delicious, satisfying meals with zero food guilt.
What it does
- Smart Virtual Fridge: Take a photo or video of your open fridge or pantry, and PortableCook analyzes your ingredients to generate instant meal concepts, smart grocery lists, and freshness alerts.
- AI Recipe Video Extraction: Paste any YouTube cooking link or upload a cooking video; the app extracts structured ingredients, exact timing, and step-by-step instructions.
- Visual Walkthrough Generator: Creates AI step-by-step image and video tutorials for tricky cooking techniques.
- Cooking Event Calendar: Schedule upcoming meals with background reminders powered by local timezone notifications.
- Tiered RevenueCat In-App Purchases: Seamlessly unlocks Pro ($25/mo) and Premium ($35/mo) plans for higher video upload limits and image generations, alongside 10- and 25-credit packs for occasional cooks.
How we built it
- Client App: Built with Flutter & Dart using Material 3 styling,
cached_network_image,file_picker, andflutter_local_notificationswith timezone handling. - Monetization Engine: Integrated
purchases_flutter(RevenueCat) using fallback lookups across package identifiers and store product IDs to prevent purchase crashes. - Backend Cloud API: Python service hosted on Microsoft Azure Web Apps (
eastus-01.azurewebsites.net) interfacing with multi-modal Gemini models.
Challenges we ran into
Handling large video file uploads on erratic mobile networks was a primary bottleneck. We resolved this by implementing client-side tier validation (limiting uploads to 5MB on Free, 15MB on Pro, and 60MB on Premium) and streaming file uploads directly through HTTP multipart requests to prevent device memory exhaustion.
Accomplishments that we're proud of
- Complete multi-modal pipeline capable of extracting structured recipe steps directly from raw cooking video files.
- Clean separation of concerns with an intuitive, one-handed UI designed for busy kitchen environments.
- Robust RevenueCat paywalls that handle both recurring monthly subscriptions and consumable credits simultaneously.
What we learned
Home cooks do not want to be judged by calorie counters. They want to reduce household food waste, save prep time, and gain the confidence to cook delicious meals with the ingredients already sitting in their kitchen.
What's next for PortableCook
- Barcode and receipt scanning for automatic pantry inventory management.
- Smart appliance integration (air fryers, smart ovens) to push temperature and timing presets directly from the app.
Built With
- flutter
- revenuecat

Log in or sign up for Devpost to join the conversation.