Rally Point

Turning scattered disaster reports into a shared picture of where help is needed.

Comment

Inspiration

I was inspired by Spidey-Tracker from the movie Spider-Man: Brand New Day. When map data can become a story, It can be amazing.

What it does

Rally Point brings public alerts, incident reports, and resident assistance requests into one workspace for Greater Dallas. An interactive map shows where events are reported, while NVIDIA Nemotron combines related evidence into summaries that help responders understand what is happening and where attention may be needed.

How we built it

I built Rally Point with React and TypeScript, using MapLibre and OpenFreeMap for the interactive map. A Python FastAPI backend collects and standardizes data from public APIs, then stores shared records in SpacetimeDB, which streams updates to connected users. NVIDIA Nemotron analyzes related reports and produces prioritized summaries with supporting evidence for human review.

Challenges we ran into

Many APIs had different data formatting so fixing that with python FastAPI was challenging but rewarding.

Accomplishments that we're proud of

Incorperating Nemotron is my favorite acheivement as I have never previously incorperated an AI in an app like this.

What we learned

I have learned to slow down while programming and not allow GenAI to to everything. This will cause major problems down the line for the app development.

What's next for Rally Point

Currently, the app only tracks the Greater Dallas area, but in the future we will expand city by city and help as many people as possible. We will also connect real world responders with this technology and get more data from the cities we are servicing.

Built With

Share this project:

Updates

Submission history