Deseeker

DeSeeker allows their customers to travel easy and cheap when their trip includes multiple destinations by checking a large arrange of possible travel scenarios, picking just the best options to show.


Client         Deseeker
Region       Barcelona, Spain
Services    Software Development
Year            2015-Present
Link            http://deseeker.com/


Features

  • User search for flights itineraries, optimized for multiple destinations trips.
  • Itinerary booking.
  • Itineraries storage and searching again based on them.

Challenges

  • The airfares API (Kiwi) calls required optimization to pull large amounts of data on a few requests, and then separate it in the single flights to analyze the scenarios.
  • Dynamic search optimization based on the Traveling Salesman Problem. It uses TSP when the number of destinations are computable and a heuristic of TSP when the number is non-computable on an acceptable time.
  • Parallel computation of best case scenario scores, using multiple threads to speed up the results calculation.

Technologies

  • Ruby on Rails 4.2
  • PostgreSQL
  • Sass, Slim and CoffeeScript
  • jQuery & Bootstrap
  • Kiwi flights connection API