PlateShare

• 

  Dashboard

• 

  Daily Log

• 

  Prep Plan

• 

  Donations

• 

  Assistant

Inspiration

We’ve worked in food service and watched perfectly good food get thrown out at the end of every shift. Meanwhile, people in our own communities go hungry. We wanted to close that gap and help small businesses waste less while feeding more people.

What it does

PlateShare helps small food businesses cook the right amount and donate what’s left. Daily log: owners record what they prepped, sold and threw out in about two minutes, or snap a photo and AI estimates the leftovers. Prep plan: recommends how much to make each day from past sales, local events and the weather forecast. Donations: finds nearby food banks, shelters and compost services, and tracks each drop-off as a ticket. Dashboard: shows money wasted, waste reduced, meals donated, CO₂ avoided and an estimated tax deduction. Assistant: answers questions using the business’s own data, like “What am I wasting the most money on?”

How we built it

We built a Next.js 16 web app with React, TypeScript, Tailwind and shadcn/ui, so it runs on any phone or laptop with no install. Server-side API routes call the Claude API for photo scanning, the assistant, prep tips and outreach emails. Open-Meteo provides free weather forecasts, and Leaflet with OpenStreetMap powers the partner map. The prep and impact numbers use simple, explainable formulas so owners can see why a number is what it is. Data is stored in a lightweight JSON store that can be swapped for a real database.

Challenges we ran into

Scoping to 3 hours: we cut features like a website builder to focus on the core loop of log, plan, donate, measure. Rethinking the donation flow: we first designed it so shelters would claim pickups, then simplified it to tickets the owner manages, since only business owners use the app. A map provider that started requiring an API key: we switched to free OpenStreetMap tiles.

Accomplishments that we’re proud of

It’s a creative, practical idea that helps small businesses save money while helping the community by reducing waste and getting meals to people who need them. In one hackathon we built a working end-to-end app: logging, AI photo estimates, forecasting, donation tracking, and an assistant that understands each business’s data.

What we learned

How to use AI APIs for image understanding and tool-using assistants. How to combine live data like weather into a forecast. How to build maps with Leaflet and OpenStreetMap. How keeping formulas simple and explainable makes the numbers easier to trust. What’s next for PlateShare A real database for secure accounts and cloud data. A mobile app so owners can photograph leftovers in one tap at the counter. Smarter forecasts: collect real customer-traffic trends to train a better model of how many customers come in each day.

Built With

• api
• open-meteo
• openstreetmap
• react
• shadcn/ui
• tailwind
• tailwind-and-shadcn/ui
• typescript