Inspiration
Every year, people waste over 1.3 billion tons of food—about one-third of all food produced globally.
Meanwhile, approximately 800 million people worldwide suffer from hunger or food insecurity.
We realized that much of this waste occurs at the household level, simply because people forget what they have, do not track expiration dates, or purchase more than they require. We wanted to create something that not only reduces waste but also enables people to make better, more sustainable food choices.
That’s how Savora was born — a smart pantry management system that allows users to track, optimize, and share their food resources.
What it does
It allows users to:
- 📦 Track inventory with quantity, category, expiry date, and nutrition info.
- 🔔 Receive smart expiry alerts before food goes bad.
- 🤖 Get AI-powered recipe suggestions to use expiring items.
How we built it
Savora was built with a focus on modern, scalable technologies and simplicity:
- Frontend: React + TypeScript + TailwindCSS for an elegant and responsive UI.
- Data Storage: Supabase as a backend database and authentication service, built on PostgreSQL, for secure and scalable pantry data management.
- AI Layer: Integrated OpenRouter’s models like
mistralai/mistral-7b-instructto power recipe suggestions.
Challenges we ran into
One of the challenges we faced was integrating Supabase for data management and authentication. Ensuring secure user sessions while maintaining pantry updates required careful schema design and permission handling.
We also had to optimize the way data was fetched and synced between the client and database to prevent redundant API calls and improve performance.
While integrating the AI layer through OpenRouter, we encountered rate limits and model response inconsistencies, which required additional validation and handling.
Accomplishments that we're proud of
- 🚀 Created a fully functional pantry tracker.
- 🧩 Integrated recipe suggestions using OpenRouter models.
- 💡 Built a project that contributes to reducing global food waste.
What we learned
Working on Savora taught us how to build a full-stack application that combines AI with food tracking and pantry management.
We learned how to design an efficient database schema in Supabase that supports dynamic pantry updates, authentication, and secure user access.
We also gained experience in frontend–backend synchronization, and improving performance through optimized data fetching.
Beyond the technical skills, we learned the importance of designing with purpose — creating software that not only works well but also promotes SDG 2 (Zero Hunger).
What's next for Savora
Future plans include:
- 🏢 Food bank connectivity to connect with local food banks and reduce food waste.
- 🧠 Improved AI predictions for shelf life and personalized nutritional recommendations.
Built With
- openrouter
- react
- supabase
- typescript

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