SkyRescue

We came up with the idea for SkyRescue after thinking about how difficult it must be to make decisions during a natural disaster, especially when there aren’t enough rescue teams to help everyone at once. We wanted to find a way to make those decisions easier to understand while exploring how AI could help. That’s how we created SkyRescue, an AI-powered disaster response simulation where users can choose between floods, wildfires, and tornadoes, explore different emergencies on an interactive map, and decide where to send limited rescue teams. We also added an AI advisor that recommends which incidents should be prioritized and an image analysis feature that lets users upload disaster photos for additional information. We built SkyRescue using React, Python, FastAPI, and OpenAI. One of the hardest parts was getting everything to work together, especially connecting the frontend to the backend and making sure the AI responses were actually helpful. We also spent time figuring out how to make the app easy to use without taking away from the seriousness of the situations. Throughout the hackathon, we learned more about building full-stack applications, working with APIs, and using AI for something beyond a basic chatbot. We’re really proud of how everything came together, especially because we were able to turn a simple idea into an interactive project that people can actually try out. If we had more time, we’d love to add more disaster scenarios, improve the AI recommendations, and make the simulations even more realistic. SkyRescue isn’t meant to replace real emergency responders, but we hope it can help people better understand the challenges they face and how technology could support them. At the end of the day, we wanted to build something that was both interesting to create and meaningful beyond the hackathon.