Devtrails Guidewire Hackathon – Blog 1

SREEJESH S CSE(CS) — Wed, 18 Mar 2026 14:43:27 +0000

RideSafe AI – Phase 1: Income Stability for Q-Commerce Workers

The Core Problem

Quick commerce (q-commerce) platforms have transformed how people receive essentials—delivering everything from groceries to daily needs within minutes. Behind this speed are delivery partners who operate in a highly unpredictable earning environment.

Unlike traditional jobs, their income is not fixed. It depends on multiple real-time factors such as order demand, weather conditions, and platform availability.

This creates a fragile system where:

Earnings fluctuate daily
Work availability is uncertain
External disruptions can completely stop income

For example, heavy rainfall or a sudden drop in orders can mean zero earnings for the day, even though expenses continue.

Why This Matters

While q-commerce is scaling rapidly, financial protection for workers in this space is still missing.

Most traditional insurance systems fail here because:

Income is inconsistent
Work patterns are dynamic
Risks are frequent but short-lived

This gap opens up the need for a system that is flexible, real-time, and tailored to the nature of q-commerce work.

Introducing RideSafe AI

To address this challenge, we built RideSafe AI—a smart income protection system designed specifically for q-commerce workers.

The idea focuses on creating a reliable safety mechanism using data and automation:

A small percentage of earnings is contributed
Contributions are maintained within a shared system
When a disruption is detected, eligible workers receive compensation

This approach enables a scalable and practical alternative to traditional protection models.

Phase 1 Objective

Our goal in Phase 1 was to validate whether this idea could work in practice.

We built a prototype that can:

Identify disruptions using real-time signals
Simulate contribution pooling
Trigger automated compensation

System Design

The system is divided into three main components:

1. Data Monitoring Layer

We integrated external APIs (such as weather data) to continuously track environmental conditions that could affect delivery activity.

2. Disruption Detection Engine

A rule-based system analyzes incoming data.

Example:

If rainfall exceeds a threshold
And delivery activity drops
A disruption event is triggered

This forms the base for future AI-driven predictions.

3. Contribution & Payout System

Workers contribute a small percentage (1–3%) of their earnings into a shared system.

When disruptions occur:

The system verifies eligibility
Compensation is automatically calculated
Payouts are triggered

Tech Stack

The system was built using a modern, scalable stack:

React Native – Mobile app for delivery partners
React (Web) – Dashboard and monitoring interface
NestJS – Backend services and API layer
PostgreSQL (PSQL) – Data storage
Prisma – ORM for efficient database management
Botpress – Conversational interface for user interaction and support

How the System Works

External data (like weather) is fetched at intervals
The backend processes this data using predefined rules
Disruptions are identified in real time
Eligible users are automatically compensated

The system is designed to operate with minimal manual intervention.

Key Outcomes

In this phase, we were able to:

Build a working end-to-end prototype
Integrate real-time external data sources
Implement automated disruption detection
Simulate contribution and payout workflows
Design a scalable and modular architecture

Challenges Faced

Some of the main challenges included:

Lack of real-world q-commerce datasets
Designing a fair contribution model for all users
Preventing misuse of automated payouts

These challenges helped us improve validation logic and system design.

What’s Next (Phase 2)

In the next phase, we plan to enhance intelligence and scalability by:

Introducing machine learning for risk prediction
Building fraud detection mechanisms
Adding region-specific analysis
Integrating with real q-commerce platforms
Dynamically adjusting contribution rates

Vision

Our aim is to build a real-time financial safety layer for q-commerce workers.

By combining:

micro-contributions
live data analysis
automated compensation

we can create a system that supports workers during uncertainty without adding complexity.

Conclusion

RideSafe AI Phase 1 proves that income protection for q-commerce workers can be built using a data-driven and automated approach.

This is just the starting point, with significant potential to scale and evolve into a real-world solution.

Let’s Collaborate

We’re open to feedback, ideas, and collaborations to take RideSafe AI further and make it production-ready.

Forem: SREEJESH S CSE(CS)