FIGGY

Inspiration

The rapid growth of the gig economy in India, driven by platforms like Swiggy, Zomato, and Amazon, has created millions of job opportunities. However, gig workers still lack financial security.

We observed that delivery partners frequently face income loss due to external disruptions such as heavy rain, floods, strikes, and curfews. Despite being active workers, they have no structured financial protection system.

This inspired us to build FIGGY, a platform that combines income protection and income optimization using AI.

What it does

FIGGY is an AI-powered parametric income insurance platform designed for gig workers.

It automatically detects disruptions and triggers compensation without requiring manual claims. The system also enhances earnings by guiding workers to high-demand zones.

Key functionalities include:

Automatic disruption detection

AI-based income prediction

Instant insurance payouts

Demand-based zone recommendations

Fraud detection using Proof-of-Work verification

How we built it

The system was developed using a modular architecture integrating AI models, backend services, and blockchain.

Frontend & Mobile: Flutter, Dart

Backend & APIs: Node.js, Express.js, REST APIs

AI / Machine Learning: Python, Scikit-learn, TensorFlow, Prophet

Databases & Caching: PostgreSQL, Redis, Pinecone (Vector DB)

Blockchain: Ethereum, Polygon

Cloud & DevOps: Docker, AWS, Google Cloud

Integrations & APIs: Weather APIs, Maps APIs, Payment Gateway APIs

AI & LLM Tools: LangChain, OpenAI

Core components include:

Demand Prediction Engine

Disruption Radar AI

Parametric Insurance Engine

Fraud Detection System

Challenges we ran into

Handling real-time disruption detection using multiple data sources

Preventing fraud such as GPS spoofing

Modeling highly variable gig worker income patterns

Integrating AI systems with backend and blockchain components

Accomplishments that we're proud of

Developed a fully automated parametric insurance workflow

Built a real-time disruption detection system

Designed a fraud-resistant Proof-of-Work verification mechanism

Integrated income protection with income optimization

Created a scalable system for gig workers

What we learned

Practical implementation of parametric insurance systems

Importance of data-driven modeling in financial applications

Designing secure and fraud-resistant systems

Integration of AI, backend, and blockchain technologies

Building solutions with both scalability and social impact

What's next for Untitled (FIGGY)

Develop a complete mobile application prototype

Integrate real-time APIs for deployment

Partner with platforms like Swiggy and Zomato

Expand to other gig sectors such as ride-sharing and logistics

Improve AI model accuracy with real-world datasets

Launch pilot testing in urban regions

Built With

• amazon-web-services
• dart
• docker
• ethereum
• express.js
• flutter
• google-cloud
• langchain
• maps-apis
• node.js
• payment-gateway-apis
• pinecone
• polygon
• postgresql
• prophet
• python
• redis
• rest-apis
• scikit-learn
• tensorflow
• weather-apis