Skip to content
 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

31 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Gigsurance - Realtime AI Driven Parametric Insurance Platform for India's Gig Economy

Presented by Team BlueTeeth

Gigsurance Banner


Breaking down Gigsurance in Four Steps

1. Who is our user really?
2. How does our AI actually work
3. How does it get built?
4. Gigsurance - Our Final Delivery

1️⃣ Who is our user really? - Tip of the ICE

Gigsurance Banner

⚠️ The Actual Problem

Introducing Team BlueTeeth - The Minds behind GigSurance

Watch our Team

One line Statement - No financial safety net for loss of working time for India's Gig Workers

One Stop Destination - GigSurance - AI Powered Dynamic & Parametric based Insurance Platform

Type of Platform - Web App

So, What type of Gig Workers are you targeting?

🧑 Revealing our Target User Persona

Gigsurance Banner

1. Our Focused Persona - Metro Cities based Gig Delivery Workers
2. Total no of active Gig Workers in Metro Cities - 25 Lakhs
3. Source link : [https://www.outlookbusiness.com/news/pay-education-demographics-what-the-numbers-say-about-indias-1-crore-gig-workers]

Our Focused Metro Cities for Gig Workers

City Estimated Gig Workers Key Risk Factors
Bangalore 4–6 lakh+ Rain, traffic, high demand
Delhi NCR 5–7 lakh+ Pollution, traffic, curfews
Mumbai 3–5 lakh+ Heavy rain, congestion
Chennai 3–4 lakh+ Flooding, cyclones
Hyderabad 2–3 lakh+ Supply disruptions
Pune 1.5–2.5 lakh+ Traffic, demand spikes
Kolkata 1–2 lakh+ Weather + infra disruptions

Our Focused Working Models

S.no Type of Model Platforms Working Time Peak hours
1. Food Delivery Zomato, Swiggy 10-12 hrs/day Lunch + Dinner Time
2. Quick Commerce Zepto, Blinkit, Swiggy Instamart 9-10 hrs/day Continuous & Seasoned Demands
3. E-Commerce Amazon, Flipkart 8-10 hrs/day Scheduled Activities

Disruptions Covering under Gigsurance

Coverage Types

✅ Covered Events

  • Extreme weather (heavy rain, cyclones, storms)
  • Flooding (delivery zone shutdowns)
  • Severe pollution (AQI thresholds)
  • Curfews and lockdowns
  • Local area strikes
  • Market crashes
  • Platform/server outages

⚠️ Not Covered

  • Health or life issues
  • Accidents
  • Vehicle repairs
  • Holidays
  • Driver out of service (without disruption)
  • Location outside registered zone

Move 1 - How We Onboard our Gig Workers

Phase 1 - The Onboarding of GigSurance Partners - Gig Workers

💡 GigSurance Mantra - 60 Sec Onboarding

❌ Not Following Traditional Insurance Model

Incident Occurs → User Files Claim → Documents Verified → Survey Done → Payment After Weeks

🛠️ How Gigsurance will work?

Deliverable - We designed onboarding as a high-speed, intelligence-driven flow that converts a gig worker into an insured user in under 60 seconds.

Our Ultra Fast Onboarding Engine - Dedicated Web App

Onboarding Flow Diagram

graph TD
    Start([Worker Opens App]) --> Phone[Phone Number Entry]
    Phone --> OTP[OTP Verification]
    OTP --> Persona[Select Work Type<br/>Food/Quick/E-commerce]
    Persona --> Location[Location Capture<br/>GPS + Zone Selection]
    Location --> Pattern[Work Pattern Input<br/>Hours/Day, Days/Week]
    Pattern --> Risk[AI Risk Calculation<br/>Location + Persona + History]
    Risk --> Plans[Plan Recommendation<br/>Basic/Standard/Premium]
    Plans --> Activate[One-Tap Activation]
    Activate --> Dashboard[Main Dashboard]
    Dashboard --> Protected([Protected!])
    
    style Start fill:#e3f2fd
    style Protected fill:#c8e6c9
    style Activate fill:#fff9c4
Loading

2️⃣ How does our AI actually work? - The Backbone under the ICE

Phase 2 - The Arrival of GigSurance Agents - Gigents

1. Gigsurance transforms insurance through parametric automation - when a qualifying event occurs, payouts trigger automatically. No claims. No paperwork. Just instant protection.

2. The Innovation: Four autonomous AI agents work 24/7 to detect disruptions, validate events, calculate fair payouts, and execute instant transfers - all without human intervention.

Let's See How our Agent Works

graph LR
    A[Disaster Event] --> B[Agent 1: Monitor]
    B --> C[Agent 2: Classify Zone]
    C --> D[Agent 3: Calculate Payout]
    D --> E[Agent 4: Instant Transfer]
    E --> F[Worker Notified]
    
    style A fill:#ff6b6b
    style E fill:#51cf66
    style F fill:#51cf66
Loading

Autonomous Agent System

Four AI agents work continuously to detect, validate, and execute payouts:

  • Agent 1: Disaster Monitor - Continuously polls weather, news, disaster APIs
  • Agent 2: Zone Classifier - Maps disruptions to specific delivery zones using NLP
  • Agent 3: Fund Allocator - Calculates individual payouts based on income baseline
  • Agent 4: Payout Executor - Instantly transfers funds via UPI

Agent System Architecture

graph TB
    subgraph "External Data Sources"
        W[Weather API]
        N[News API]
        D[Disaster API]
        A[AQI API]
        T[Traffic API]
    end
    
    subgraph "Agent Workflow"
        A1[Agent 1: Disaster Monitor<br/>Polls APIs every 5-15 min<br/>Calculates confidence score]
        A2[Agent 2: Zone Classifier<br/>NLP processing<br/>Maps to delivery zones<br/>Estimates severity]
        A3[Agent 3: Fund Allocator<br/>Queries active policies<br/>Calculates payouts<br/>Fraud validation]
        A4[Agent 4: Payout Executor<br/>UPI transfers<br/>Notifications<br/>Audit logging]
    end
    
    subgraph "Data Layer"
        DB[(PostgreSQL)]
        VDB[(Pinecone Vector DB)]
    end
    
    W --> A1
    N --> A1
    D --> A1
    A --> A1
    T --> A1
    
    A1 -->|Threshold Exceeded| A2
    A2 --> A3
    A3 --> A4
    
    A2 --> VDB
    A3 --> DB
    A4 --> DB
    
    style A1 fill:#4dabf7
    style A2 fill:#4dabf7
    style A3 fill:#4dabf7
    style A4 fill:#4dabf7
Loading

Real-Time Fraud Detection

Multi-layered security system:

  • GPS spoofing detection through movement pattern analysis
  • Behavioral verification (active hours, delivery routes, speed consistency)
  • Multi-device triangulation prevention
  • Trust score system (0-100) updated in real-time

Dynamic Pricing Model

Premium adjusts based on:

  • Location Risk Score - Historical disruption frequency in zone
  • Persona Type - Food delivery vs Quick commerce vs E-commerce
  • Work Pattern - Hours per day, days per week
  • Real-Time Conditions - Current weather trends, seasonal patterns

Predictive Intelligence

AI analyzes historical data to suggest optimal premium plans before disruptions occur, helping workers make informed decisions.


3️⃣ How Does it GigSurance get build? - The Architecture

Phase 3 - The Brick by Brick Solid Architecture of GigSurance

GigSurance Deliverables

Primary Focus: Core Platform + 60-Second Onboarding + Autonomous Payout System

What We're Building in Phase 1:

The foundation of Gigsurance - a fully functional parametric insurance platform that demonstrates the core value proposition: instant, automated protection for gig workers.

Web Application for Gig Workers

Gig Worker Onboarding in 60 Seconds

The fastest insurance activation in the market:

  • Phone verification with OTP
  • Persona selection (Food Delivery / Quick Commerce / E-commerce)
  • Location capture (GPS + manual zone selection)
  • Work pattern input (hours/day, days/week)
  • AI-powered plan recommendation (Basic / Standard / Premium)
  • One-tap insurance activation

Main Dashboard Features:

  • Live insurance status and coverage details
  • Real-time location tracking
  • Trust score display
  • Payment history
  • Weekly renewal management
  • AI chatbot for support

Admin Dashboard (Next.js) - Dedicated Web Application

Event Monitoring & Management

Complete control center for disruption management:

  • Real-time disruption event feed
  • Zone visualization with affected workers
  • Manual event triggering capabilities
  • Payout approval queue
  • Worker management and verification
  • Analytics and reporting

Backend System (FastAPI)

Core Infrastructure:

  • RESTful API for mobile and admin dashboards
  • Agent orchestration system
  • Risk calculation engine
  • Dynamic pricing algorithm
  • Fraud detection pipeline
  • Payment integration (Razorpay)

Technology Stack

Layer Technology
Mobile Web App (Angular JS)
Admin Dashboard Next.js 14 (TypeScript)
Backend FastAPI (Python)
AI/ML LangChain + OpenAI
Database PostgreSQL + Pinecone (Vector DB)
APIs Weather, News, Disaster, AQI, Traffic
Payments Razorpay (UPI)
Notifications SMS/WhatsApp API

GigSurance - Real Time Scenario

Scenario 1: Heavy Rainfall

Event: Rainfall exceeds 120mm in Chennai
Trigger: Agent 1 detects threshold breach
Action: Agent 2 maps affected zones → Agent 3 calculates payouts → Agent 4 transfers ₹500 to each worker
Time: Complete in under 10 minutes

Scenario 2: Delivery Zone Shutdown

Event: Local strike or curfew closes delivery zones
Trigger: News API + Traffic API detect disruption
Action: Automatic payout based on estimated income loss percentage
Time: Instant after validation buffer

Scenario 3: Cyclone Warning

Event: Wind speed > 80km/h detected
Trigger: Weather API + Disaster API confirmation
Action: Proactive alert sent to workers + automatic coverage activation
Time: Pre-emptive protection before impact


Competitive Advantages

Traditional Insurance Gigsurance
Weeks-long processing Instant payouts
Manual claims filing Fully automated
Document verification GPS-based validation
Designed for salaried Built for gig workers
Fixed pricing Dynamic, location-based
Reactive coverage Predictive alerts

Innovation Highlights

1. Parametric Automation

Unlike traditional insurance, payouts are triggered by objective data (rainfall mm, wind speed km/h) rather than subjective claims. This eliminates fraud, speeds up processing, and ensures fairness.

2. Zone-Level Precision

We don't just detect city-wide events - our AI maps disruptions to specific delivery zones, ensuring only affected workers receive payouts.

3. Rolling Weekly Cycles

Each worker has an individualized 7-day cycle starting from their signup time, not a fixed Monday-Sunday schedule. This accommodates part-time workers and flexible schedules.

4. Cooldown Period

New users have a 6-hour cooldown before coverage activates, preventing exploitation during active disruptions.

5. Location Intelligence

System automatically detects if a worker has been in a different city for 24+ hours and prompts location update, ensuring accurate coverage and pricing.


Getting Started

For Gig Workers

  1. DNavigate to GigSurance Web App
  2. Complete 60-second onboarding
  3. Select your work type and location
  4. Choose your insurance plan
  5. Start earning with protection

For Administrators

  1. Access the admin dashboard
  2. Monitor real-time events
  3. Review payout approvals
  4. Manage workers and policies
  5. Analyze system performance

Contact & Support

For inquiries or support:


License

[MIT] - [2026] Gigsurance


Submitted at GuideWire DevTrails Univeristy Hackathons by Team BlueTeeth

❤️ Built for India's Gig Workers - GigSurance

❤️ Taking a Moment to Show gratitude to Guidewire for organising these type of hackathons

About

GigSurance - AI-enabled parametric insurance platform that safeguards gig workers against income loss caused by external disruptions such as extreme weather or environmental conditions.

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages