Hello I am Case Zumbrum and I will be submitting into the Bookings category. My submission for the Square Unboxed Hackathon is a booking prediction algorithm. My project uses the Square Bookings API to access all bookings made throughout the course of a year at a particular business. I also used used the Customer API to connect customer ID’s to names, which was used when displaying data.
For this project, I decided to use a neural network to analyze the data. The neural network is trained using the booking data from each user, with the algorithm being fit to the next appointment for that user. I realized that the neural network functioned worse if a user had few past appointments, to combat this I created a second algorithm that uses the average time between appointments to estimate the date of the next appointment. This algorithm is used when a user has less than 5 appointments, which I found to be a good cutoff point for the neural network.
My primary challenge throughout this project was how to use the data from the Bookings API to make useful predictions with a neural network. This primarily involved how to convert data to a form usable by the neural network, and then convert it back into a human-readable format. I am proud of the precision that the neural network showed when making predictions, and I learned a great deal about how to manipulate data to be usable by many programs throughout a project.
I used Python for both analytical algorithms and PHP to access the predictions and display it in a browser, and I hosted the project on an Apache web server. To access my submission, go to casezumbrum.hopto.org and use “square” as both the username and password. The data is divided into two groups, the simple algorithm and the neural network. The names of users are on the left, and the date of their predicted next appointment is on the right. This project could be used for any square business, but it is currently set up to use my testing sandbox.
Please contact me at casezumbrum@gmail.com if you are unable to access anything!
Youtube Link: https://youtu.be/CtcycRkFa0g
Log in or sign up for Devpost to join the conversation.