Second Serve

Inspiration

Good food goes to waste while people in our communities struggle to access nutritious meals. Restaurants, supermarkets, and farms may have surplus food to donate, but getting it to the right place is only part of the challenge. Community kitchens also need to decide what to cook, which ingredients to use first, and how to accommodate different dietary needs.

We created Second Serve around a simple idea: rescuing food should lead to usable meals. By connecting food donations with smarter meal planning, we want to help communities make the most of every ingredient.

What it does

Second Serve connects food donors with community kitchens and helps turn available ingredients into meal plans.

The project brings together:

• Surplus food listings: Restaurants, supermarkets, and farmers can share food available for donation.
• Ingredient prioritization: Kitchens can identify ingredients that need attention first using reported dates and storage information.
• AI-assisted recipes: Meal suggestions help kitchens use the ingredients they have.
• Dietary considerations: Meal planning takes reported allergies and dietary restrictions into account, with human review before preparation.
• Flexible portions: Kitchens can adjust recipes and serving quantities to make better use of available food.

Our broader vision includes coordinating pickups, exploring more efficient delivery routes, and directing suitable scraps toward approved agricultural or composting partners.

How I built it

I approached Second Serve as a web app with two connected workflows: listing surplus food and planning how to use it.

I broke the idea into a donor-facing flow for entering food details and a kitchen-facing flow for reviewing ingredients and planning meals. The AI component focuses on suggesting recipes from available ingredients, while structured information such as quantities, dietary restrictions, and reported use-by dates provides context.

For the hackathon, I focused on the central journey: surplus food → donation → kitchen inventory → meal suggestion.

Challenges I ran into

The biggest challenge was narrowing a large idea into a realistic project as a solo participant at my first hackathon. Donation matching, inventory management, recipes, transportation, and waste diversion could each become a project on their own.

Another challenge was designing around uncertainty. An AI-generated recipe cannot establish whether donated food is safe, and missing allergen information cannot be treated as confirmation that an ingredient is allergen-free. This made clear information and human review essential parts of the concept.

Meal planning also involves competing priorities: using urgent ingredients, accommodating dietary needs, and producing useful quantities of food.

Accomplishments that I'm proud of

I'm proud of developing a project that connects food rescue with the practical decisions kitchens face after a donation arrives.

Second Serve goes beyond finding a recipient for surplus food. It asks how that food can become meals people can actually eat, while helping kitchens use more of what they receive.

What I learned

I learned that AI is most useful when it supports a clear workflow and works within explicit constraints. Generating a recipe is only one step; making it useful requires quantities, ingredient availability, dietary information, and human judgment.

I also learned the importance of controlling scope. As a first-time solo hacker, defining one coherent user journey helped me turn a broad social-impact idea into a more focused project.

What's next for Second Serve

Next, I want to test the workflow with community kitchens and food donors to understand what would make it useful in daily operations.

Future improvements include donation reservations, pickup coordination, route planning, clearer ingredient tracking, and impact reporting based on completed donations. I also want to explore Toronto food-access data to help identify underserved areas.

The goal is simple: give good food a second chance to serve the community.

Built With

• css
• fastapi
• git
• google
• next.js
• postgresql
• python
• react
• tailwind