Conversational Travel AI

Inspiration

The goal of this project is to bring the power of Azure Cognitive Services and Voice Search to the Travel Industry. There are a number of problems that our project aims to solve:

Lack of voice and conversational capabilities in Travel products- it is estimated that 50 percent of all searches will be voice searches by 2020 according to ComScore
Conservative design and lack of personalization
Need for multiple channels and consistent user experience
Poor accessibility for people with disabilities

What it does

Conversational AI system for flight and hotel bookings
End-users can use chat and voice to find and book trips
Automated service that saves costs and time
Combining the power of Conversational AI with traditional Travel systems
Available on multiple channels e.g. web, mobile app, social networks, etc.
Accessible for people with disabilities

Features

Conversational AI communication via voice and chat
Search and book flights via connection with the Amadeus APIs
Search and book hotels via connection with the Amadeus APIs
Conversational AI mobile app
Answer FAQs about travel

How we built it

The project is heavily based on several Azure Cognitive Services.

Speech Services - this service is used to generate the voice/speech output of the assistant
LUIS.ai - this service is used for parsing natural language queries/messages; we recognize the intents and entities of each message
QnA Maker - this service is used for answering frequently asked questions; this service uses Azure Search internally

High-level architecture

The source code of the project is split into two main parts:

client - contains all front-end code, implementing the user interface server - contains all back-end code, implementing responses, flows, controllers, API integrations, etc.

The front-end of the project is implemented with Vue.js. The back-end is implemented with Node.js.