insureMe

• 

  Insurance Preferences

• 

  Welcome screen - select your location and dates

• 

  Result - What insurance should you get!

• 

  Activity Recommendations Selection -1

• 

  Activity Recommendations Selection - 2

Inspiration

We are a team from Singapore and when we booked our flights to San Francisco, we realised travel insurance was a hassle as we were provided with many policies with different features such as loss of limbs, loss of eyesight which were not what we needed for our trip. We figured out that individuals want to customise their journey of purchasing their travel insurance and should not be provided with a fixed benefit illustration.

We are a team from Singapore and when we booked our flights to San Francisco, we realized travel insurance was a hassle as we were provided with many policies with different features such as loss of limbs, loss of eyesight which were not what we needed for our trip. We figured out that individuals want to customise their journey of purchasing their travel insurance and should not be provided with a fixed benefit illustration. However, the process of doing so is really hard and complicated.

What it does

InsureMe allows individuals to customise their travel insurance policy based on the features of the policy. We used Amadueus API to fetch text on the location data of consumers and provide them with personalised recommendations on travel insurance required for the specific location. We allow consumers to select the features of insurance products they want to ensure they only pay for the coverage they opt in for. Based on the location data, we can categorise the types of activity in the area and provide insurers with the risk level of consumer on their trip to make better underwriting decision. Travel insurance can be purchased in 3 steps:

• Select your location you are travelling to
• Select the features of your policy
• Select the type of activities you are participating in

How I built it

Building this solution first involved creating a mockup using proto.io to ensure our team was on the same page. Following a discussion on suitable implementation tools, we decided to choose the versatile React Native as our framework of choice for building the application. In addition to this, we also built a containerized microservice using python and docker that enables us to do some backend tag probability scoring for use in the application.

Challenges I ran into

All of us are backend engineers and had no experience with frontend. In this hackathon, we learnt React Native! Amadeus API has a limit of 10 results of each location, which limited the number of cards we could show.

Accomplishments that I'm proud of

We managed to get the mock machine learning recommender microservice to serve the main app.

What I learned

Most of our team has never built and application before. So our primary takeaway was technical. We got our first exposure to the expo client and also figured out how to integrate APIs to match our needs. We also gained familiarity with version control software as we actually established a GitHub team and had to learn how to manage a local, personal remote and a team remote repositories.

What's next for insureMe - Recommendation System

We want to improve our recommendation system using machine learning. With customer preferences, and features of the places from Amadeus API, we can train a deep learning classifier to get the probability of the travel activity being relevant. With that, we can refer people to locations and earn referral fees from ticket prices, increasing our revenue.

Built With

• python
• react