Inspiration
Flight booking can be a time-consuming task. Not only you need to browse different flight booking sites, you must also compare different flights based on various factors, such as price, nearby airports, flight schedule, airline preference. Therefore, our group decides to solve this problem with an easy-to-use web application with ChatGPT-powered interactive chatbot to complete this process in 30 seconds or less!
What it does
- Create an airline chatbox that when the users type in their cirtiria, it going to look up the suitable flight for them.
- Extract users' input to JSON file using OpenAI.
- Insert the JSON file to Amadeus API to generate the matching results to another JSON file.
- Use OpenAI to convert the result JSON file back to NLP and output the answer to the users.
Challenges we ran into
- Convert the input cities to their airport code.
- Convert the different input dates to a uniform format
- Handle the front-end input to make an understandable JSON file in the back-end for further generation.
Accomplishments that we're proud of
- Create a functioning chatbox website with appealing minimalistic UI
- Manage large flights dataset in order to give an appropriate response.
- Implement a feature that transfer speech to text input.
What we learned
- We learned to work a lot with API and NLP models.
- We learned how to host the Flask Python microservice on Google Cloud Platform using Google Cloud Run.
Demo
Watch our demo video here: https://youtu.be/fDNwpJHFoFQ
Connect with us on Linkedin!
Conan Nguyen: https://www.linkedin.com/in/nhat-nguyen-conan/
Hung Bui: https://www.linkedin.com/in/hungqbui/


Log in or sign up for Devpost to join the conversation.