EmbrAlert is a full-stack wildfire detection, prevention, and community alert application designed for the diverse San Jose area and beyond. It provides real-time wildfire risk assessments, weather and air quality updates, AI-driven smoke detection, multilingual chatbot support, and live camera wildfire detection.
🛠️ This project was built during SJHacks, a 24-hour hackathon hosted in San Jose, California. Our goal was to create a proactive solution to help communities detect and prevent wildfires before they spread.
- 🔥 Real-time wildfire risk prediction using a lightweight RNN model
- 🌎 Live weather and air quality data dashboard
- 📸 Upload images or use live camera for instant smoke detection
- 💬 Multilingual chat system powered by a custom RAG pipeline (supports six languages)
- 🎙️ Voice and text input capabilities
- 🌐 Optimized for both web and mobile devices
- Node.js and npm installed
- Python 3.x installed
- AstraDB (or another vector database access)
-
Open a terminal and navigate to the
/serverdirectory:cd server -
Install dependencies:
npm install
-
Start the server:
npm run dev
-
Server will run at:
http://localhost:3001
-
Open another terminal and navigate to the
/clientdirectory:cd client -
Install dependencies:
npm install npm install lucide-react
-
Start the client:
npm run dev
-
Client will run at:
http://localhost:5173
- Access the app at http://localhost:5173
- The app will automatically connect to the backend server running at http://localhost:3001
- Public launch with emergency authority integration
- Expand the RNN model with additional wildfire datasets
- Add support for additional languages
- Deploy scalable versions for wildfire-prone areas globally
MIT License. Feel free to fork and build upon EmbrAlert!
Sarthak Sethi, Edwin Yue, Samson Xu, Tanzil Ahmed