Skip to content

2k33cse992574/GigShield

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

4 Commits
 
 
 
 
 
 
 
 

Repository files navigation

GigShield — AI-Powered Parametric Insurance

Designed for India's Gig Economy | Hackathon Project


Executive Summary

GigShield (Full-Stack Event-Driven Platform) | Node.js, React.js, Express, MongoDB Live: guidewire-app.onrender.com

  • Architected a MERN-stack platform with an automated trigger-evaluation engine; designed indexed MongoDB schemas across 5+ entity types, reducing trigger lookup latency as evaluation volume scaled.
  • Engineered a real-time trigger engine monitoring disruption events and developed a dynamic premium calculator utilizing an AI-driven risk-scoring formula to ensure sustainable pricing.
  • Built a high-fidelity, futuristic 3D interface to significantly enhance user trust and accessibility for non-technical gig partners.

🏗 Microservices Architecture

To handle real-time data ingestion, seamless user interactions, and heavy machine learning processing, GigShield uses a modern microservices architecture:

1. frontend/ (React.js + 3D Interface)

The client application used by Delivery Partners and Admins.

  • Built with React.js (Vite).
  • Integrates React Three Fiber and Three.js for the futuristic 3D UI gateway.
  • Contains the legacy HTML/JS dashboard under public/ for backward compatibility.
  • To run: cd frontend && npm run dev

2. backend/ (Node.js API Gateway)

The core backend serving as the orchestrator for user data and API requests.

  • Built with Node.js, Express, and MongoDB (Mongoose).
  • Manages User Auth, Claims Storage, and routes ML requests.
  • To run: cd backend && node server.js

3. ml_service/ (Python AI Engine)

Dedicated service for Machine Learning and complex calculations.

  • Built with Python, FastAPI, and scikit-learn.
  • Contains the Isolation Forest ML model for fraud detection and risk scoring.
  • Processes the real-time trigger engine (Weather APIs, AQI, Traffic).
  • To run: cd ml_service && pip install -r requirements.txt && uvicorn api:app --reload

⚙️ How It Works (End-to-End Workflow)

Worker Registers on GigShield
          ↓
AI Risk Scoring (weather, AQI, zone risk, work pattern)
          ↓
Weekly Premium Calculated → WI × R × 0.05
          ↓
Trigger Engine polls APIs every 30 minutes
          ↓
Disruption Detected (rain / heat / AQI / flood / curfew)
          ↓
Multi-Signal Validation → Movement Check → Active Policy Check
          ↓
Fraud Detection Model (Isolation Forest) runs instantly
          ↓
If Approved → Payout credited to UPI instantly!

🚀 Running the Project Locally

To run the full stack locally, you need to spin up all three services:

  1. Start the ML Service (Port 8000)

    cd ml_service
    pip install -r requirements.txt
    python main.py  # Or run `uvicorn api:app --reload`
  2. Start the Node.js Backend (Port 5000)

    cd backend
    npm install
    node server.js
  3. Start the React Frontend (Port 5173)

    cd frontend
    npm install
    npm run dev

Developed for DEVTrails 2026

About

No description, website, or topics provided.

Resources

Stars

1 star

Watchers

0 watching

Forks

Releases

No releases published

Packages

 
 
 

Contributors