SIP Profile: Wyatt Hansen
An auto DJ for Spotify and its creator.
The following is a transcript of a short interview conducted with Wyatt Hansen (’23) about his Senior Independent Project — Playlist-Assist. Billed as “an auto DJ for Spotify,” this program analyzes users’ Spotify music and modifies their current queue to ensure that songs flow smoothly from one track to the next.
Q. What factors have influenced your pursuit of learning to program? Moreover, how have you gone about teaching yourself coding?
A. I taught myself how to code in late October 2022. I’ve definitely started later than a lot of people, but I’ve found a serious passion for it, and that’s been enough to keep me learning. I finally made the decision to teach myself because I needed to create a video game for a business I’m working on. To answer the second question, I’ve found project-based learning to be absolutely essential. I never really taught myself out of a book or off a website. I’d just start a project I was interested in and fix specific problems as I encountered them.
Q. When did you first get the idea of creating a program like Playlist-Assist? What was your main motivation in building it?
A. I got the idea for Playlist Assist from a website I found a while ago called The Eternal Jukebox. It’s basically the same idea as Playlist-Assist, but instead of transitioning between songs, The Eternal Jukebox jumps around within a single song, creating an infinite loop. I basically wanted to build an Eternal Jukebox for multiple songs, and that led me to Playlist-Assist.
Q. What has been the most rewarding part for you in creating this program? What has been the most frustrating part?
A. I’ve found that the most frustrating parts of projects are often the most rewarding. I think certainly my favorite part of Playlist Assist is the way it reorganizes playlists. It tries to create a “path of least resistance” within the playlist so that when it starts thinking about jumps, they’ll be easier to find. Designing that system was probably the most difficult part of the project, but it’s contributed the most to making transitions sound smoother.
Q. What is your next major goal in the development of Playlist-Assist?
A. Right now, I’m just trying to make it more consistent and reliable. It’s capable of making amazing transitions, but it also fails a lot. Over the next few weeks I’ll probably be working on other projects, but I’m hoping that sometime over the summer, I’ll be able to improve the beat-finding algorithm and increase its reliability.
Q. Do you have any other projects that you are currently working on? If not, what would you like to build or develop in the future?
A. I’m working on a couple projects currently, but I think the most notable is TOS Chat. It’s a custom AI chatbot that can answer questions about company agreements like privacy policies, community guidelines, and terms of service. I’ve been busy finishing up Playlist-Assist. But now that I’m done with that and school is almost over, I’ll be working on TOS Chat a lot more. It works right now, but over the next couple weeks, I’m hoping to add a fancier landing page and iron out some bugs.
Q. Finally, where do you plan on going to school? What do you plan on studying?
A. Next year, I’ll be studying computer science at Colorado State University, but I’m going to try to transfer back to UT after freshman year.
You can find out more about Wyatt’s project here!