Several insurance firms have tried to break into the activity-specific insurance plan market with limited success. Here, by the term activity-specific insurance market, we are specifically referring to those short tailed insurance plans that target potentially risky short term activities. Since these plans are short term, the insurance company doesn't have to wait for a long time to know whether or not they have made money or not. Since the entire market seemed like a large beast to tackle head-on, we thought of specifically targeting those travellers who have a higher likelihood of indulging in high risk activities such as adventure sports and high altitude mountaineering.
Next, we had to make the user realize his/her need for such an insurance. Every time a long weekend is around the corner, several people, especially college going students like ourselves, begin looking for places for a potential outing. Since there are a huge number of websites offering everything from recommendations to reviews to bookings, it becomes a cumbersome (and often tiresome) process to sift through potential locations, trying to see which place fits our expectations of what is ideal.
Travel Jarvis is a single platform where you can learn more about holiday activities and destinations that are tailor-made for you, and also optionally buy a reasonably priced, customized insurance plan.
What it does
In a nutshell, Travel Jarvis is a personality based travel risk assessment and short tailed insurance recommendation system.
At a high level, it infers the user’s personality from their Twitter profile and provides a custom list of attractions. Next, the user can view the risk associated with each of these attractions and decide which ones that he would potentially visit. (Some risks that are shown are criminal risk, safety risks and precipitation risks. These risks are calculated from data that is obtained from APIs.) We also use Bluemix's Trade-off Analytics API to maximize the reward to risk ratio. Then, based on the selections, a collection of insurance plans (from the database of the insurance policy aggregator) are suggested to the user. If the user chooses an insurance plan that isn't being offered yet, the insurance aggregator is notified, so that they may take a decision on whether or not to offer that policy in the future or not.
How we built it
We started by using the twitter handle of the person to get insights about his personality and then giving a list of a attractions suitable for the person in the desired place. Then we use several API’s for getting data related to precipitation, crime, safety etc in that location. This data was used to create the risk profile of a tourist attraction/ adventure activity for the days he wishes to visit. Then we use IBM Watson’s Trade-Off Analytics API to help the user choose between the various attractions based on the risk and his interest. Then the user can choose several places that he wishes to visit and then he can get recommendations for insurance that he or she can buy or wish to buy.
Challenges we ran into
Using the trade-off analytics API because the sample app uses a static json file from the file system but using a custom JSON fetched from the back-end took us a lot of time. Apart from this, trying to come up with a user interface that matches the high expectation of the users these days.
Accomplishments that we're proud of
We are proud of the fact that we got a application up and running by the end of the hackathon. We jumped into untested waters and continuously ran into bottlenecks one after the other. But with a strongly willed team and many experts to guide us, we were eventually able to put our thoughts into a working application.
What we learned
We worked on the IBM’s Bluemix Platform and used some very powerful API’s that it provided . We learned that design thinking is really important and makes a huge difference, even for a one day long project. By thinking of the user from the beginning, our entire development perspective took a different turn.
We also learned tremendously from both the domain experts from Swiss Re and the technology experts from IBM. Overall, the hackathon was a rewarding experience for all of us.
What's next for Travel Jarvis
In future Travel Jarvis would act as a platform where the Insurance Companies come and realize the demand for short tailed insurances as posted by the users. They get an opportunity to get introduced to new products based on demand. It can create a whole new market of personalised, activity specific insurances, a market that insurance companies have tried for a long time to break into.