Inspiration
COVID-19 and other outbreaks demonstrated how slow, reactive resource allocation can cost lives. We wanted to build a tool that models disease spread realistically and helps decision-makers act fast.
What it does
Cura simulates epidemics in real time using a tile-based model and optimizes resource deployment (vaccines, medical staff, supplies) to minimize spread. Users can visualize infections across a map and test “what-if” scenarios interactively.
How we built it
- Backend: Python + Flask API serving simulation data
- Frontend: React + TypeScript with a Google Earth–style interactive map
- Data: US Census tracts, population density, climate, healthcare access
- Algorithms: Two-level stochastic SIR model, spatial network analysis using Queen contiguity, and resource allocation optimizer
Challenges we ran into
- Integrating real census data with geospatial boundaries for 83,000+ tracts
- Modeling both local spread and long-distance jumps (airports) realistically
- Ensuring real-time frontend performance while running complex simulations
Accomplishments that we're proud of
- Full real-time epidemic simulation on a map with live infection counts
- Implemented a resource optimizer that prioritizes high-risk areas like schools and low-income regions
- Created a scalable architecture capable of handling US-wide data efficiently
What we learned
- How to combine geospatial analysis, stochastic modeling, and frontend visualization
- Practical challenges of scaling simulations for thousands of nodes while keeping the UI responsive
What's next for Cura
- Add multi-pathogen simulations with different transmission models
- Implement reinforcement learning to optimize resource allocation policies automatically
- Enable collaborative scenario planning and exportable policy briefs for real-world use
Built With
- flask
- gsap
- mapbox
- python
- radix
- react
- tailwind
- typescript
- vite
Log in or sign up for Devpost to join the conversation.