Around three million Canadians live in poverty. For these citizens, one major setback can mean the difference between putting food on the table and going hungry. Our goal was to create an app to streamline the microinsurance process for those who may need it the most, allowing them to have more security in their lives.
What it does
InsurApp uses image recognition and machine learning services powered by AWS to identify and appraise the user's assets, which we then use to recommend them insurance plans that are the best bang for their buck.
Read more about the app: https://github.com/ilPikachu/automated-micro-insurance/blob/master/README.md
How we built it
InsurApp is comprised of an Android app, our matching API, and the item database.
Frontend Android: This app utilizes Jetpack navigation component, retrofit2 and okhttp3.
Backend: Flask, MySQL Database, AWS Textract, AWS Rekognition, Keras Neural Network.
Read more about frontend app: https://github.com/ilPikachu/automated-micro-insurance/blob/master/README.md
Read more about backend: https://github.com/Piceptron/MicroinsuranceRecommender
Challenges we ran into
The biggest challenges we ran into were establishing connection between the app and the API, sending the image data, and developing the recommendation algorithm.
When sending the image data, we ran into performance issues converting the BitMap into binary data and sending that to the API.
Accomplishments that we're proud of
We are proud to have brought our idea to life within the timeframe and despite the challenges that we had. In addition, it was our first time using Amazon Web Services so we are very happy that we were able to successfully utilize it to its fullest in our project.
What we learned
We learned how to set up and utilize Amazon Web Services
What's next for InsurApp
InsurApp still has room to grow, such as allowing users to submit claims for their insured items through the app.
Team Members: Bob Bao, Hao Cong Su, Michael Dai, Jarrod Servilla