The entire airline booking experience so far has been web driven and lacks personalization. The ease of taking help from an AI would make the booking experience much more pleasurable and efficient. We also know about possible flight disruptions based on past data but they are seldom provided to customers during a generic web booking. Knowing the customer can help provide personalized suggestions and avoid issues like having a passenger book a flight that requires transit visa.
What it does
Our app is a virtual assistant that can book flights based on the user profile and can also help with ease of booking the best possible flight along with a quick and seamless experience.
How we built it
We have used AWS marketplace to use ML/AI applications that can be readily integrated with Amazon Lex. We look at customer profile from the customer database as well as use merchandising solutions to help personalize offerings.
Challenges we ran into
We have plans to create our own Sagemaker module to use past flight delays to help predict future delays for a flight/airport on a specific date. This is still in development and we hope to integrate it to make the application useful for passengers who prefer being on-time over other metrics like cost, transit requirements, etc.
Accomplishments that we're proud of
We were able to use the AWS marketplace to build the skeleton of the entire application within a week. Our reference architecture could be built and deployed quickly, as it was very easy to build enterprise grade applications using the AWS APIs.
What we learned
We learned that the opportunities are endless once we decide to build applications on the AWS stack. We hope to improve this product and hopefully make it available on the AWS marketplace.
What's next for Virtual Assistant to book Flights & plan travel Itinerary
We want to expand it to cover the entire booking and check-in experience that can replace a typical Internet Booking Engine (IBE).