- Project Name: ZapSure
- Short Project Description: A ML based insurance app which makes claiming insurance a task of seconds
- Team ID:HX019
- Team Name:Psifrous
- Team Members: Naman Bansal @psifrous, Swapnil Panwala @panwalas, Rushil Goomer @R-Goomer
- Repository Link: https://github.com/psifrous/Zapsure
- Labels: influence the masses,android, ML, image processing, AWS, CNN, Deep Learning
An insurance agent in your pocket which will make insurance claims as easy as clicking a selfie, now you dont need to wait couple of weeks for getting your health/car insurance, this app will assess the damage of your car based on a photo clicked by you and then by using multiple CNN models we will predict the amount to be claimed by the user, in this app we took a model which checks for the car in the image, check for damage, check for area of damage, check for severity of the damage. The project was trained on an ec2 server on AWS with about a dataset of 2500 images of different angles and severity.
Current Opportunity 🚀 There has not been a major disruption in the insurance industry since decades and has been a major money hauling industry, hence to bring disruption Zapsure will help the citizens to be connected to a common source to claim their car insurance.
Future Opportunity ⏲️ As technology moves ahead, so will we Zapsure will be soon including Health Insurance where in users could give in their XRAY scans, LAB reports etc. and that will assess the health claim in according to the database given by the respective hospital.
#Model Overview
CNN Model on 3 different scenarios:
🔥This project was made for HaXplore hackathon held at IIT BHU