Dibs by jetBlue

Inspiration

At first, we were motivated to use JetBlue's dataset to generate models to predict flight patterns in the future. When we found out the specific hack that we were being asked to implement, we realized that it would be an excellent opportunity to create a web application that serves as both a simple interface for consumers and an effective sales technique to fill distressed packages.

What it does

Dibs displays a neatly-organized grid of potential destinations for the wanderlust traveler. Beautiful photographs taken from relevant Instagram posts that illustrate the scenery of the destination in question are overlaid with the name of the location. If the consumer is interested, they click and are provided with the bare minimum information: the ground covered by their flight, a photograph of the resort they will be staying at, and a stream of photos that display the activities that take place at this location. If they decide that the time and trip are right, they click "Book Now", which takes them to JetBlue's website to book their getaway.

How I built it

Since this is a mockup, we carefully selected the most appealing images we could find to serve as front-page eye-catchers. These components are generated by react.js. The followup page uses Google Maps' API to illustrate the flight path on a map, while the streaming images are animated by react.js (these images are also curated; more on that in the next section). Flask was used as a web framework and Heroku was used to host the application.

Challenges I ran into

The most significant challenges we had to deal with were API issues. For instance, Instagram's API is neither thorough nor reliable: posts older than 7 days cannot be fetched, and direct media retrieval is only possible through a pipeline of API calls. As a result, we were only able to retrieve posts that were only geographically relevant: content was random and often inappropriate. After recognizing these issues, we attempted to move forward and use Twitter's API to retrieve the same information in analogous ways. This proved to be challenging by virtue of Twitter's strict limitations of use: we were unable to retrieve the information required before getting locked out.

Accomplishments that I'm proud of

The design is sleek, simple, and arguably quite effective. If marketed towards those who tend to take vacations on a whim, they are provided with only as much as they need to make their decision: pictures, encouragement, and scope of travel. Individuals who are likely to impulsively travel are not interested in the particulars, such as (but not limited to) price, specific activities/itinerary, and departure time.

What I learned

We learned a lot about cross-compatibility of API's and how things don't always work out as well as you'd like them to. Less-experienced members of the team were provided an opportunity to learn about react.js.

What's next for Dibs

Dibs boasts an incredibly simple interface that could be implemented in a very short amount of time. If refined and developed by means of substantial research, Dibs could plausibly reduce the loss incurred by JetBlue when getaway packages are left unclaimed.