Disaster Alert & Safety Intelligence Bot

Inspiration

Climate change is increasing the frequency of floods, cyclones, heatwaves, and extreme weather events. Many communities receive alerts too late or lack clear guidance on what actions to take. During recent climate events in India, misinformation and delayed communication caused panic and preventable damage.

This inspired us to build an AI-powered system that not only alerts people but also provides intelligent, personalized safety guidance in real time.

What We Built

We developed Disaster Alert & Safety Intelligence Bot, an AI-driven platform that:

• Monitors real-time climate and disaster data
• Predicts disaster risk using machine learning
• Provides location-based alerts
• Suggests personalized safety instructions
• Detects and verifies misinformation
• Identifies nearby shelters and emergency services

The system transforms raw climate data into actionable safety intelligence.

How We Built It

Backend

• Python with FastAPI
• REST API integrations for weather and disaster data
• Risk prediction using basic regression models

AI Engine

• NLP-based alert generation
• Context-aware safety recommendations
• Misinformation classification logic

Frontend

• React.js with Tailwind CSS
• Interactive dashboard
• Map-based visualization for nearby shelters

Database

• MongoDB / PostgreSQL for storing alerts and user data

Key Innovation

Unlike traditional alert systems, our solution:

• Predicts risk instead of just reporting it
• Provides personalized safety guidance
• Integrates misinformation detection
• Focuses on climate resilience and preparedness

Impact

• Improves disaster preparedness
• Reduces panic through verified information
• Enhances emergency response efficiency
• Supports climate action through technology

Challenges We Faced

• Integrating multiple real-time APIs
• Handling inconsistent disaster data formats
• Designing accurate risk scoring models
• Ensuring alerts remain simple and understandable

Despite these challenges, we successfully built a scalable and intelligent prototype that can evolve into a government-integrated disaster management system.

Built With

• fastapi
• machine-learning-(regression)
• mongodb
• natural-language-processing
• openstreetmap-api
• python
• react.js
• rest
• tailwind-css
• weather-api