Skip to content

codewithRakz/Team_Apex-Innovators_DEVTrails_2026

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

3 Commits
 
 

Repository files navigation

GigShield AI

AI-Powered Parametric Insurance for Food Delivery Partners


Team Details

  • Team Name: Apex Innovators
  • Hackathon: Guidewire DEVTrails 2026
  • Phase: Ideation & Foundation

Persona

We are building for Food Delivery Partners (Swiggy/Zomato) operating in urban Indian cities such as Chennai and Bangalore.

These workers rely on daily deliveries for income and are highly affected by environmental disruptions like rain, heat, and pollution, which directly reduce their working hours and earnings.


Problem Statement

Food delivery partners in India frequently lose income due to external disruptions such as heavy rainfall, extreme heat, and severe air pollution. These conditions can reduce or completely stop their working hours, leading to 20–30% weekly income loss. Currently, there is no insurance solution focused on protecting their income, leaving them financially vulnerable.


💡 Our Solution: GigShield AI

GigShield AI is an AI-powered parametric insurance platform that provides income protection to delivery workers by automatically detecting disruptions and triggering payouts.

  • No manual claims
  • No paperwork
  • Fully automated payouts
  • Weekly subscription model

Parametric Trigger Scenarios

Payouts are triggered automatically when conditions are met:

Trigger Condition Payout
Heavy Rain Rainfall > 50 mm/day ₹400
Extreme Heat Temperature > 42°C ₹300
Severe Pollution AQI > 300 ₹250

These conditions directly impact delivery operations and cause income loss.


💸 Weekly Pricing Model

The platform uses a weekly premium model aligned with gig worker earnings.

Risk Level Weekly Premium
Low Risk ₹20/week
Medium Risk ₹35/week
High Risk ₹50/week

Risk Factors:

  • Location-based weather history
  • Frequency of disruptions
  • Area-specific risk patterns

AI/ML Strategy

1. Dynamic Premium Calculation

AI models analyze location and historical disruption data to assign a personalized weekly premium.

2. Income Loss Prediction

The system estimates potential income loss based on working hours affected by disruptions.

3. Fraud Detection

Machine learning models detect anomalies and suspicious claim behavior using multi-signal validation.


🔐 Adversarial Defense & Anti-Spoofing Strategy

1. Differentiation

The system uses a Multi-Signal Trust Score instead of relying only on GPS.

Signal Real Worker Spoofer
Movement Active Static
GPS Pattern Continuous Sudden jumps
App Activity Accepts deliveries No activity
Network Type Mobile data WiFi
Weather Match Matches location Mismatch

2. Data Used

The system analyzes multiple data points:

  • Device sensors (accelerometer, gyroscope)
  • GPS movement patterns
  • Network type and IP validation
  • Delivery activity (simulated)
  • Weather API comparison
  • Nearby user behavior (cluster validation)

3. UX Balance

To ensure fairness:

  • Low Risk: Instant payout
  • Medium Risk: Quick verification
  • High Risk: Delayed review

Users are informed transparently without blocking genuine claims.


Anti-Collusion Detection

The system detects coordinated fraud using:

  • Same IP clusters
  • Simultaneous claims
  • Repeated patterns

🔄 System Workflow

  1. User registers and enters location and income
  2. AI calculates risk score
  3. Weekly premium is assigned
  4. System monitors real-time environmental data
  5. Trigger condition is detected
  6. Claim is automatically initiated
  7. Fraud detection is performed
  8. Payout is processed instantly

Integration Plan

  • Weather APIs (or mock data)
  • AQI APIs (or mock data)
  • Delivery activity simulation
  • Payment gateway (Razorpay test mode / mock UPI)

📊 Dashboard Features

Worker Dashboard

  • Active policy status
  • Risk level
  • Earnings protected
  • Claim history

Admin Dashboard

  • Total users
  • Claims triggered
  • Fraud alerts
  • Risk analytics

🛠️ Tech Stack

  • Frontend: HTML, CSS, JavaScript
  • Backend: Python (Flask / FastAPI)
  • AI/ML: scikit-learn
  • Database: Firebase / MongoDB
  • APIs: Weather + AQI

Development Plan

Phase 1

  • Idea and architecture

Phase 2

  • Registration
  • Premium calculation
  • Trigger system
  • Auto payout

Phase 3

  • Fraud detection
  • Dashboard
  • Optimization

Conclusion

GigShield AI provides a fully automated, AI-powered insurance solution tailored for India’s gig economy. By focusing on income protection, weekly pricing, and fraud-resistant architecture, the platform ensures financial stability for delivery partners during external disruptions.

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors