-
-
Home Page
-
Login Page
-
Empty Form
-
Completed Form
-
Output Card Stack
Inspiration
Vinay Gupta, one of our teammates, recently went to Bangkok, Thailand over this past summer, and is an avid traveler. However, he had trouble planning for the trip and what places to visit as there are a large variety. Noticing a pattern in difficulty of finding anything from niche local spots to lively tourist attractions; With collective repetitive past troubles as well, we all decided to take matters in our own hands and build a tool that will give us the most important activities that reflect our personality, and maximize our vacation experience.
What it does
Explorer.AI takes in basic parameters of the user's desired vacation in that it takes the location, the month of visiting, number of days visiting, the pace of the activities and vacation, number of people traveling with, and the level of budget. After the user fills out this short form, the software will take in theses parameters and run in through an AI model to output an day-by-day itinerary curated for the specific user. The output is neatly outputted in a card stack format, highlighting detailed schedule for the user to explore.
How we built it
We built the back-end/machine learning processor in Python using several machine learning APIs such as sGPT to process and personalize an extremely detailed and accurate output. Next, we constructed a database using the SupaBase framework implementing a REST API to save specific user ID's, authenticate users, create accounts, save personal trips, trip metadata. Next, we began constructing the front-end using React.JS, JavaScript, HTML and Tailwind CSS. We constructed the landing page, where users are welcomed by a futuristic dashboard with the option to log-in, create an account, or generate a trip. We also incorporated email verification for account setup. The next page is a multi-typed form which receives user input to pass into the back-end python script. This process also saves user data and the python output in the database, organized by user and time sent. To display the output, we implemented a Flask framework in Python which directly communicated between the front-end and back-end. With this data, we parsed, formatted, and uploaded the final result with an animated card stack using React. Each page has dynamic 4K video content in the background, elevating the professionalism and engagement for the user.
Challenges we ran into
One of the first challenges we ran through the entire time was learning and programming simultaneously. Our group didn't have much experience in using the SupaBase, React, and tailwind CSS. Not only this was challenging as we kept running into errors, it was time consuming as well. The other main problem we ran into was developing the front end of our website. Our group was mostly experienced in back-end so developing a website was a struggle as we kept running into syntax errors. We even asked the mentors at midnight for help, but even they couldn't solve it.
Accomplishments that we're proud of
One of the main accomplishments we are proud of is returning a card style itinerary which each card displaying the day number of the trip and all the activities planned sectioned out by "morning", "afternoon", and "evening". This was our main purpose of the tool so by completing this successfully, made us very proud. Another accomplishment was successfully connecting the database to the front end and back end of the program. Before this project, our team members had minimal experience in databases and API calling. So after a large learning curve and hours of our time, we successfully connected it.
What we learned
We learned many programming languages and frameworks that we haven't even used before. Through this project, we learned and became familiar with Supabase, flask, REST API's, html, tailwind css, and javascript. Apart from the technical learning, our group learned collaborative skills as working on a single project for nearly 24 hours made sure we all communicated clearly and had the same vision for each sprint and end of the project.
What's next for Explorer.AI
The goal is to expand the user interface so that users can view their personalized profile; the profile would contain a travel card that highlights the user’s past trips and allows them to see, on a map, all the places they have visited so far. Additionally, we would like Explorer AI to have a suggestions and preferences tab so that users can input what activities or events that they prefer or want to avoid. The suggestions will be asked after the user has read all of the data that the program outputs; it will allow users to make edits to their specific travel plan taking into account their personalized parameters. Lastly, we would love to expand the services offered by including both rental car and hotel plans which could help users create a complete plan of their vacation.
Log in or sign up for Devpost to join the conversation.