Inspiration

I noticed that restaurants, cafes and shops throw out a lot of food and goods at the end of the day, and regular people throw out things they no longer need. At the same time, people who could use those things live right nearby.

What it is

ResQShare lets a donor (a person, a restaurant or a shop) post surplus food or items, and people and organizations nearby can pick them up. The AI is responsible for distribution. It decides who a donation should go to, taking into account distance, the recipient's real needs, dietary limits, urgency, family size and perishability (in that case it routes the food to organizations instead of individual people). The AI also writes the listing from a photo, so the donor barely has to type anything, and it estimates the contribution of each completed donation.

How I built it

Android app: Kotlin and Jetpack Compose.
Backend: Java and Spring Boot, REST API.
AI: I used ChatGPT 5.4 over the API, through the Replicate provider.
Hosting: the backend runs in a Docker container on my own server (Hetzner). It is running right now, so anyone can download the app and use it.

What I learned

Working with maps was new to me, so that users can drop points showing where they are or where the items they give away are located.

Using AI to build the project

About 70% of the code was written by hand, and the other 30% was written by ChatGPT. I built the project architecture myself, wrote many of the functions and worked on the app's design (inspired by Duolingo). The more routine, boilerplate functions I delegated to ChatGPT.