Inspiration
Most people don't know about the different types of insurance policies that StateFarm has to offer. In fact, before our research, we thought insurance was synonymous with auto insurance. We wanted to create an app to help people understand the different types of insurances that are suitable for them.
What it does
Our Mobile app asks the customer a serious of questions and matches them to the types of insurances suitable for them. It then takes the customer to a chat bot to help them understand what the insurance does as well as additional information about the insurance.
How we built it
We built the front-end using React Native, and used Flask as the back-end. In addition, we created a chat bot using PyTorch. The chat bot includes a Natural Language Processing algorithm that first separates a sentence into single words, excludes special characters, then stems the word by removing the prefix and suffix. The word is then sent to a neural network to find hidden patterns (trained thousands of iterations) and returns the output to the user. Finally, We made our application scalable so StateFarm can easily integrate any new products into our App using existing components.
Challenges we ran into
A huge hurdle we ran into was connecting the chat bot with the front-end since neither of us had written API calls before.
Accomplishments that we're proud of
We were able to learn React-Native and Pytorch from scratch, and use them to create an entire application in under 24 hours with no prior knowledge of mobile development. We learned our limits when it comes to learning and coding as well as team work and communication.
What we learned
We were able to learn the fundamentals of React-Native and Pytorch in addition to creating API calls for information from the backend. We successfully set up Google VPC to route APIs calls from our backend to our frontend.
What's next for InsuraFIND
In the future, we want to directly connect our chat bot to the app for better customer experience. We are already able to make API calls using flask and Google VPC. We would also like to integrate a machine learning algorithm to give better suggestions to customers.
Log in or sign up for Devpost to join the conversation.