DealDish Story 🍽️

Inspiration

We were tired of trying to piece together weekly ads while balling on a budget. It is so time consuming to come up with healthy recipes with ingredients that are on sale. In today's economy, with rising grocery costs and food waste from unplanned purchases, we needed a better solution that could help us make smarter shopping decisions while still eating well.

What it does

DealDish uses screenshots of weekly grocery ads to curate healthy meals based on food preferences/allergies, difficulty level, and cuisine. The app:

Automatically processes images of grocery store deals
Generates personalized recipes based on sale items
Considers dietary restrictions and allergies
Helps reduce food waste through smart meal planning
Saves money by focusing on ingredients currently on sale
Stores and organizes your favorite recipes

How we built it

We started out on a FigJam coming up with problems we were facing. From there, we started working on the app coding and branding. Our technical implementation includes:

Modern Tech Stack: React with TypeScript, Tailwind CSS, and shadcn-ui for a beautiful, responsive interface
Powerful Backend: Supabase for real-time database and authentication
AI Integration: OpenRouter with Google's Gemini model for intelligent recipe generation
Type Safety: End-to-end TypeScript implementation for robust code quality
Mobile First: Fully responsive design that works seamlessly on all devices
Smart Search: Full-text search capabilities for recipes and cuisine tags

Challenges we ran into

Some errors and illogical recipe suggestions (i.e. strawberry/grape salsa (?))
Balancing between AI creativity and practical recipe suggestions
Ensuring accurate deal recognition from various grocery store ad formats
Maintaining type safety across the entire application
Optimizing the mobile experience while keeping full functionality

Accomplishments that we're proud of

We're just excited to make something we'd actually use! We believe that this will actually save us valuable time and money on a weekly basis!

What we learned

It's not about glamour, it's about practicality. The core functionlity is the key and we shouldn't get too carried away with the UI or features. Key technical learnings include: