Official Submission for DEVTrails 2026 — Seed Phase 1
Protecting 8M+ delivery partners from income loss caused by extreme weather, AQI, and curfews — powered by real-time environmental APIs and AI fraud detection.
Architecture • Features • Installation • Demo • Setup Guide
India's gig workers lose up to 30% of their weekly earnings due to events beyond their control: torrential rain, hazardous air pollution (AQI > 400), and municipal curfews. Traditional insurance requires manual filing, documentation, and weeks of processing—proving impossible for a worker missing a single day's wage.
GigShield AI is a zero-touch parametric insurance platform.
- Auto-Triggered Claims: Our engine monitors WAQI and OpenWeatherMap APIs. If a threshold is hit, claims are created automatically.
- Instant Payouts: Verified claims are disbursed instantly via RazorpayX to the worker's UPI ID.
- Affordable Weekly Premiums: Micropayments (₹30-₹120) tailored to gig pay cycles.
- AI Sentinel: An Isolation Forest AI model scores every claim for fraud in real-time.
- One-Click Protection: Buy weekly coverage for specific disruptions (Rain, Heat, AQI).
- Live Razorpay SDK: Real-time premium collection with UPI, Card, and Netbanking support.
- Virtual Wallet: Track every premium paid and claim received in a glassmorphism ledger.
- Risk Heatmap: Live Leaflet.js map showing disruption risks across major Indian metros.
- Trigger Engine Control: "Force Env Scan" button to manually sync with weather satellites.
- Fraud Command Center: AI-flagged claims are listed for human-in-the-loop verification.
- Data Export: One-click CSV export for all claims and financial records.
- Frontend: React 19, Vite, Chart.js, Leaflet, Framer Motion.
- Backend: Node.js 22, Express, Razorpay SDK, PostgreSQL 16.
- AI Microservices: Python 3.12, FastAPI, Scikit-learn (Random Forest & Isolation Forest).
- Infrastructure: Docker Compose (7-service microservice mesh), Redis 7 (Caching).
GigShield AI is fully containerized. Deploy the entire stack with one command:
git clone https://github.com/dkv204p/GigShield-AI.git
cd GigShield-AI
docker compose up --build -d- Frontend Dashboard: http://localhost
- Admin Login:
9999999999/admin123 - Worker Login:
9876543210/worker123
To test the "Parametric Trigger" logic immediately:
- Log in as Admin.
- Go to Claims Monitor.
- Click "Force Env Scan".
- The system will fetch real-time weather/AQI for major metros. If values are hazardous, you will see claims populate instantly in the dashboard!
(For detailed troubleshooting and manual setup, see SETUP.md)
This project is licensed under the MIT License — see the LICENSE.md file for details.
Empowering the backbone of India's digital economy.
Made with ❤️ for the Gig Community.