Paper Plane

Inspiration

We came up with the inspiration for our project while searching for flight tickets to get from Zurich to New York and realized that it was cheaper to buy the exact same tickets from the Finnish version of the Skyscanner website. It turns out that the online travel agencies (OTAs) choose to price their tickets according to the geographical market. This means there are price discrepancies for the same flight tickets. This hack arbitrages these price discrepancies.

What it does

With Paper Plane users can query and visualize the cheapest flights from Skyscanner without having to check manually over all geographical markets. Basically, we traverse the list of prices for all the flights for all the markets and display only the top cheapest ones.

How we built it

We built PaperPlane mainly in Python, Flask and using web technologies. We made extensive use of the Skyscanner API. We hosted it on Heroku.

Code:

link

Challenges we ran into

The biggest challenge that we faced was the communication between the frontend and backend. Performing a query on the Skyscanner API takes a lot of time so we had to find a way to stream the results such that they continuously appear to the users.

Accomplishments that we're proud of

We are proud that we've built a fully functional prototype which can already help people save money on their flights. We used AWS to run our application in parallel and gather a big dataset of flight pricing. We then run analyses on this and extract insightful information. On average, an inter-continental flight might have an 80eur pricing discrepancy depending on which market it is bought from. We also noted a weak correlation between the market the ticket is bought in, with the distance the market is from the departure location or destination location. This means that there is no way to know in advance which markets to look at, proving the usefulness of PaperPlane.

What we learned

During the hackathon, we learned a about web development using Python and HTML/JS and also got experience developing an API backend server using Flask.