Skip to content

Repository files navigation

GigShield

AI-Powered Parametric Insurance for India's Gig Delivery Workers

Protecting Income. Not Just Lives.

An AI-driven, fully automated parametric insurance platform that protects gig workers against income loss caused by heavy rain, floods, extreme heat, air pollution, traffic shutdowns, government restrictions, and platform outages.

Built for Guidewire DEVTrails 2026

Status Version License AI-Powered Blockchain InsurTech


Vision

Millions of India's gig workers lose income every year because of events beyond their control — heavy rain, floods, extreme heat, air pollution, curfews, platform outages. Current insurance products protect vehicles, health, or life. They do not protect lost earnings.

GigShield changes that. Workers receive compensation automatically whenever verified external disruptions prevent them from earning. No paperwork. No claim forms. No waiting.


Problem

India has ~12 million platform-based gig delivery workers. They earn per order — miss a shift, earn nothing. A single rainy day can wipe out 20–30% of weekly income. No insurance product covers this gap.

Traditional Insurance GigShield
File a claim manually Zero-touch — auto-triggered
Submit proof of loss Objective data = proof
Wait 3–15 days for settlement Payout in under 15 minutes
Requires employment proof No employment proof needed
Annual, one-size-fits-all Weekly, AI-personalized premium

Solution

GigShield combines AI, parametric insurance, real-time weather intelligence, fraud detection, blockchain transparency, and instant UPI payouts into one autonomous income protection platform.

How it works:

External Event (Rain / AQI / Curfew / Outage)
        ↓
Trigger Engine detects + dual-source verifies
        ↓
Policy match → Fraud screening (14 signals)
        ↓
Payout calculated → UPI transfer initiated
        ↓
Blockchain logged → Rider notified
   (under 15 minutes, zero manual steps)

Key Features

Feature Description
AI Risk Assessment Personalized weekly premiums per rider, zone, platform, season
Automatic Trigger Detection Real-time monitoring — Rain, AQI, Heat, Curfew, Outage
Zero-Touch Claims No manual submission — auto-detected and auto-paid
AI Fraud Detection 14 signals: GPS, physics, device, behavior, network collusion
Income Bridge Advance payout on high-confidence triggers before full verification
Micro-Shift Insurance Hourly coverage for individual work sessions
Blockchain Transparency Immutable on-chain record of every trigger, claim, and payout
Loyalty Pool Community mutual pool — 10% of unclaimed premiums redistributed
Analytics Dashboards Rider, Admin, Executive, Fraud dashboards with real-time data

Live Demo

Service URL
Frontend https://gigshield-frontend-4ad7.onrender.com
Backend API https://gigshield-backend-4wmh.onrender.com/health
ML Service https://gigshield-ml-swb3.onrender.com/health

Free tier — first load may take 30–60 seconds (cold start)

Demo login: Phone 9821000001 → OTP from backend logs → Ravi Kumar's dashboard

Admin login: Phone 9800000001


Demo Accounts

Phone Rider Scenario
9821000001 Ravi Kumar Happy path — GREEN claim, paid, blockchain-logged
9821000002 Priya Sharma Income Bridge advance → reconciled after selfie
9821000003 Amit Singh RED fraud — mock location detected — appeal pending
9821000004 Sunita Devi Gold loyalty tier, 11-week streak
9821000005 Mohammed Farhan KYC in progress, referred by Sunita
9821000006 Deepak Yadav Active Micro-Shift policy
9821000009 Lakshmi Nair Income Bridge → clawback (verification timeout)
9821000010 Rajesh Patel Fresh empty-state account
9800000001 Admin Full admin + fraud + executive dashboard

OTP appears in backend terminal: [DEV] OTP for ***0001: 482913


Tech Stack

Layer Technologies
Frontend React 18, Vite, Redux Toolkit, TailwindCSS, Socket.IO, Framer Motion
Backend Node.js, Express, MongoDB Atlas, Redis (Upstash), Bull queues, Socket.IO
ML Service Python, FastAPI, scikit-learn, XGBoost, LightGBM
Blockchain Solidity, Hardhat, Ethereum Sepolia, Ethers.js
Infrastructure Render (PaaS), MongoDB Atlas, Upstash Redis, GitHub Actions

Architecture

┌─────────────────────────────────────────────────┐
│                 GIGSHIELD PLATFORM               │
├───────────────┬────────────────┬─────────────────┤
│   FRONTEND    │    BACKEND     │   ML SERVICE    │
│  React + Vite │ Node + Express │ Python FastAPI  │
│  Redux + WS   │ Bull + Cron    │ XGBoost + IF    │
└───────┬───────┴───────┬────────┴────────┬────────┘
        │               │                 │
        ▼               ▼                 ▼
   Render CDN    MongoDB Atlas      Render Docker
                 Upstash Redis
                 Blockchain (Sepolia)

Project Structure

gigshield/
├── backend/        Node.js + Express API, workers, cron jobs
├── frontend/       React + Vite SPA
├── ml-service/     Python FastAPI — ML risk & fraud models
├── blockchain/     Solidity smart contracts (Hardhat)
├── nginx/          Reverse proxy config
├── Docs/           Full project documentation (30+ files)
├── SETUP.md        → Local development guide
├── DEPLOYMENT.md   → Production deploy guide (Render)
└── .env.example    → Environment configuration template

Getting Started

Local Development

See SETUP.md — complete guide with account creation, API keys, and troubleshooting.

# 1. Clone + install
git clone https://github.com/Ambivert01/AI-Parametric-Insurance-for-India-s-Gig-Workers.git
cd AI-Parametric-Insurance-for-India-s-Gig-Workers
cd backend && npm install && cd ../frontend && npm install && cd ..

# 2. Setup environment
cp .env.example .env
ln -s ../.env backend/.env
# Edit .env — fill MONGO_URI, JWT secrets, ENCRYPTION_KEY

# 3. Seed demo data
cd backend && npm run seed

# 4. Start services (3 terminals)
cd backend && npm run dev                    # API → :5000
cd frontend && npm run dev                   # UI  → :3000
cd ml-service && source venv/bin/activate && uvicorn main:app --reload --port 8000

Production Deployment

See DEPLOYMENT.md — Render deployment guide with MongoDB Atlas, Upstash Redis, all 3 services.


Security

  • JWT authentication with refresh tokens
  • Phone + OTP login (no passwords)
  • Role-based access control (Rider / Admin / Insurer)
  • AES-256 encryption for bank/UPI details
  • Rate limiting (per endpoint + edge)
  • Fraud scoring on every claim (14 signals)
  • Audit logging on all mutations
  • Blockchain immutability for payouts

Future Roadmap

  • Mobile apps (iOS + Android)
  • IoT sensor network integration
  • WhatsApp AI assistant
  • Satellite weather data
  • Embedded insurance APIs for platforms
  • DAO governance for loyalty pool
  • International expansion

Documentation

Full technical and business documentation in Docs/:

File Contents
PROJECT_DOCUMENTATION.md Complete business + technical master doc
01_SYSTEM_ARCHITECTURE.md System architecture
06_AI_ML_SYSTEM.md ML models and risk engine
07_TRIGGER_ENGINE.md Trigger detection system
08_FRAUD_ENGINE.md Fraud detection system
30_MASTER_INDEX.md Full docs index

Development Principles

  • Clean Architecture + SOLID Principles
  • Event-Driven Communication (Bull queues + Socket.IO)
  • AI-Assisted Decisions (ML scoring on every claim)
  • Zero-Trust Security (fail-closed everywhere)
  • Documentation-First Development

GigShield — Protecting Every Delivery. Protecting Every Income.

Built with ❤️ for India's 12 million gig workers

Guidewire DEVTrails 2026

About

Next-generation parametric insurance platform combining AI risk scoring, real-time trigger detection, blockchain transparency, and zero-touch payouts to protect gig workers from income loss in a fast-changing world.

Topics

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages