Laser Eyes Bot is a Twitter bot that lets you automatically add laser eyes to your Twitter profile photo with one tweet. When triggered, the bot will pull the tweeter’s profile picture and run it through a machine learning algorithm to detect faces and eyes. Any detected eyes will be turned into laser eyes and the resulting photo will be tweeted out! The bot has made over 30,000 impressions on Twitter and dished out over 70 laser eyes photos. I built this with Node.js, Python, and OpenCV and hosted it on a Raspberry Pi.
Zeitplan is an all in one productivity tool with reminders, daily scheduling, logs, and a Pomodoro timer for efficiency. I cofounded this project with classmates at CPP and lead development with the MEAN technology stack. We built it as a progressive web app with a mobile first design and we deployed it on AWS.
For my final project in my machine learning elective, I chose to build a machine learning model to predict beauty in order to automate online dating. Using Python and Keras, I trained a machine learning model that ranks facial beauty using the SCUT-FBP500 dataset from the South China University of Technology. Using Python and an unofficial Tinder API, my program detected faces from Tinder images and fed them into my convolutional neural network. If the model gave a good score, the bot would “swipe right” or like the person. I ran this bot overnight and it swiped on thousands of people, quickly netting me over 100 likes. I have since decommisioned this bot because I found it immoral, but I learned a ton building it and it proved to be a short term solution to automating online dating.
Vibe is a cross-platform social media app for students that helps them plan impromptu local activities with their friends. I built this project with a couple friends in my early college days using React Native, Algolia Search, Node.js, and Firebase.
Throughout college, I participated in multiple hackathons. Aside from the networking benefits, hackathons served as a great way to improve my software development and presentation skills. Building a software product in one weekend can be quite hectic!
I developed many Android apps while learning how to code during my junior and senior years of high school. I published all of my apps in excitement to share with my friends. These apps remain on the store today and some sport a healthy number of users. My favorites are A to B - Schedule Your Tasks, Flappy Monkey, and the Texas Hold’em Expected Value Calculator.