During disasters people struggle to:
- Find safe shelters
- Access resources
- Contact family
- Reach emergency services
- Navigate dangerous areas
- Receive verified information
Existing systems are fragmented. Raksha AI unifies everything into a single emergency survival platform.
Raksha AI acts as an AI-powered survival copilot.
It helps citizens:
- Find shelters
- Locate resources
- Generate survival plans
- Connect with family
- Request help
- Navigate safely
- Access offline emergency guidance
| Feature | Description | Screenshot |
|---|---|---|
| AI Survival Copilot | Adaptive AI assistant providing step-by-step guidance tailored to the user's emotional state and emergency severity. |  |
| Disaster Map | Real-time map displaying current disaster zones, safe routes, and nearby critical infrastructure. |  |
| Shelter Finder | Instant location of verified, safe shelters based on the user's proximity and route viability. | |
| Resource Discovery | Locate and access essential emergency resources, crowdsourced and updated by the community. |  |
| Emergency SOS Network | One-tap panic SOS system that instantly alerts emergency contacts and authorities with GPS coordinates. |  |
| Family Safety Tracker | Centralized hub to verify the safety and live location of family members. |  |
| Community Rescue Network | Decentralized mesh of nearby volunteers and helpers who can assist before official responders arrive. |  |
| Offline Survival Guides | Cacheable, easy-to-read emergency response protocols for offline access during network outages. |  |
| Emergency Alerts | Live intelligence, news updates, and government broadcasting for active disaster events. |  |
Click to view walkthrough
graph TD
A[User reports emergency] --> B[AI analyzes situation]
B --> C[Shelters/resources identified]
C --> D[Safe route generated]
D --> E[SOS sent if required]
E --> F[Family updated]
F --> G[Survival guidance provided]
| Domain | Technologies |
|---|---|
| Frontend | Next.js, React, TypeScript, TailwindCSS, Framer Motion |
| AI Layer | Google Gemini |
| State Management | Zustand |
| Testing | Playwright |
graph TD
U[User] --> R[Raksha AI]
R --> EA[Emergency Analysis]
EA --> RE[Resource Engine]
EA --> SE[Shelter Engine]
EA --> SOS[SOS Engine]
EA --> SG[Survival Guidance]
- Problem: Rising water levels, disconnected routes, power outages.
- Raksha AI: Guides to highest ground, alerts nearby boats/rescuers, offline survival steps.
- Problem: Collapsed buildings, panic, aftershocks.
- Raksha AI: Panic-mode guidance ("Drop, Cover, Hold On"), immediate SOS broadcast.
- Problem: Severe winds, flying debris, communication breakdown.
- Raksha AI: Advanced warning alerts, shelter routing before impact.
- Problem: Rapidly spreading fire, smoke inhalation, unpredictable paths.
- Raksha AI: AI mapping of safe evacuation routes away from wind trajectory.
- Problem: Remote locations, lack of medical help, severe trauma.
- Raksha AI: One-tap GPS dispatch to 112 and contacts, first aid steps for bystanders.
- Problem: Trapped individuals, dust, low visibility.
- Raksha AI: Periodic low-power SOS pinging, battery conservation mode.
- Problem: Cardiac arrest, choking, sudden illness.
- Raksha AI: Step-by-step CPR/Heimlich guidance matched to the user's stress level.
- Problem: Heatstroke, dehydration, vulnerable populations.
- Raksha AI: Reminders, cooling station locator, symptom checker.
- ✔ Faster emergency response: Seconds saved in critical moments.
- ✔ Better survival planning: Actionable steps replacing panic.
- ✔ Shelter discovery: Immediate routing to safety.
- ✔ Family coordination: Peace of mind through automated check-ins.
- ✔ Community collaboration: Decentralized aid bypassing overwhelmed hotlines.
- ✔ Offline accessibility: Crucial guidance when cell towers fall.
- Mobile support: Fully responsive design tested on 320px+ viewports.
- Keyboard navigation: Full tab support and semantic HTML.
- Responsive design: Fluid layouts adapting to all screen orientations.
- Offline capability: Essential survival guides cached locally.
- Multi-language support: Auto-detects and supports Hindi, English, and Hinglish.
Ensure you have Node.js and npm installed.
# Clone the repository
git clone https://github.com/hackindia-team/ai-agents-hackathon-2026-sigmoid.git
# Navigate to the app directory
cd ai-agents-hackathon-2026-sigmoid/raksha-app
# Install dependencies
npm install
# Set up environment variables
# Copy .env.example to .env.local and add your Gemini API key
cp .env.example .env.local
# Run the development server
npm run devai-agents-hackathon-2026-sigmoid/
├── raksha-app/
│ ├── src/
│ │ ├── app/ # Next.js app router pages
│ │ ├── components/ # Reusable React components
│ │ ├── hooks/ # Custom React hooks
│ │ ├── lib/ # Utility functions and API clients
│ │ └── store/ # Zustand state management
│ ├── tests/ # Playwright E2E tests
│ └── public/ # Static assets
├── docs/
│ └── screenshots/ # Application screenshots
└── README.md
The platform includes a robust suite of E2E tests covering all critical paths.
# Build the application
npm run build
# Run linting
npm run lint
# Execute Playwright tests
npx playwright testRunning 9 tests using 1 worker
ok 1 [chromium] › tests\screenshots.spec.ts:8:7 › Raksha AI V2 - Screenshots & E2E › Capture Home Page (16.5s)
ok 2 [chromium] › tests\screenshots.spec.ts:25:7 › Raksha AI V2 - Screenshots & E2E › Capture Disaster Map (2.0m)
ok 3 [chromium] › tests\screenshots.spec.ts:40:7 › Raksha AI V2 - Screenshots & E2E › Capture Alerts (33.5s)
ok 4 [chromium] › tests\screenshots.spec.ts:47:7 › Raksha AI V2 - Screenshots & E2E › Capture Resources (11.0s)
ok 5 [chromium] › tests\screenshots.spec.ts:54:7 › Raksha AI V2 - Screenshots & E2E › Capture Family & SOS (33.3s)
ok 6 [chromium] › tests\screenshots.spec.ts:70:7 › Raksha AI V2 - Screenshots & E2E › Capture Community (1.3m)
ok 7 [chromium] › tests\screenshots.spec.ts:77:7 › Raksha AI V2 - Screenshots & E2E › Capture Survival Guide (59.0s)
ok 8 [chromium] › tests\screenshots.spec.ts:84:7 › Raksha AI V2 - Screenshots & E2E › Capture Emergency Assistant (54.0s)
ok 9 [chromium] › tests\screenshots.spec.ts:97:7 › Raksha AI V2 - Screenshots & E2E › Capture Impact Page (20.5s)
- Build status: Passing
- Lighthouse: Designed for 90+ across all metrics
- Optimization: Image optimization via Next.js, code splitting, edge-ready API routes.
- Real-time disaster feeds (API integrations)
- Satellite integration for completely offline messaging
- NGO & Government portal integrations
- Offline mesh networking via Bluetooth/WiFi direct
- AI predictive risk analysis based on weather patterns
- Multi-country and multi-regional language support
We welcome contributions from the community. Please follow these steps:
- Fork the repository
- Create a feature branch (
git checkout -b feature/AmazingFeature) - Commit your changes (
git commit -m 'Add some AmazingFeature') - Push to the branch (
git push origin feature/AmazingFeature) - Open a Pull Request
Distributed under the MIT License. See LICENSE for more information.
Built by: Prakhar Goel
For: AI Agents Hackathon 2026
Mission: Using AI to improve disaster preparedness, response, and survival.