Data scientist and software developer with 9+ years of entrepreneurial experience who excels at machine learning and data storytelling, I enjoy the crossovers of data and music and see data as a tool for effective strategizing and problem-solving. Seeking to leverage my expertise in diverse applications of data science.
.png)
Alexaurus Recs is a Spotify recommendation engine deployed on Heroku using a Flask web app framework. Alexaurus Recs aims to improve upon Spotify song recommendations by comparing the audio-features of a user's most-listened-to songs with those of base recommendations from Spotify. Under the hood, cosine similiarity is calculated between each song's audio-features vector and a list of recommendations is returned. I use Spotipy, a python wrapper for Spotify's web API, to retrieve user-data.
Please visit my GitHub Repo for more information.
This project explores the current state of text generation utilizing GPT2 and aitextgen. I created an interactive chatbot trained on dialogue from the hit sitcom, Seinfeld, to create an entertaining and functional AI. Given a prompt, this AI produces coherent and believable Seinfeld-esque dialogue.
Additionally, the chatbot has a built in character predictor. Given a person's chat dialogue, the bot will predict which main character you sound most like. Lastly, the chatbot will recommend episodes based on how you interact. Based on transfer-learned sentence level embeddings, it calculates which episodes "sound" most like you. Each recommendation provides an option to hear the plot and a direct link to watch the show. Below is a demonstration.
Using data collected by the New York Times, this project aims to explore the spread of COVID-19 in regards to mask sentiment and predict daily case "regime" changes using a unsupervised Hidden Markov model. My team developed a suite of applications that help visualize the spread of COVID-19 and plot daily case change along with predicted "regime" change. The hope is that a user can check the website to see if we have entered a new "phase" of virus spread as well as plot potential Super Spreader events to compare timelines.
Utilizing Reddit's API, I scraped over 30,000 posts from the "Machine Learning" and "Artificial Intelligence" sub-Reddits. Using advanced NLP techniques and various classification models from scikit-learn, I was able to predict with 85% accuracy which sub-Reddit a post came from using a Random Forest Classifier. Recently, I have updated my project with Tensorflow's most current NLP classification framework.
This project details my take on the classic, Ames, IA housing dataset. Using a classic ridge regression, custom feature engineering and hyperparameter tuning via grid-search, I placed second in a cohort Kaggle competition predicting the sale price of homes in Ames, IA. From this model, I was able to glean which home characteristics and features were most commonly correlated with higher sale price and subsequently make recommendations to homeowners to increase resale value.
I develop, deploy and maintain monetization and churn models for popular video games like Rainbow Six Siege and Assassin's Creed.
Create consumer facing UI and connect to Salesforce backend.
Immersed in a 13 week, 500+ hour course focused on bleeding-edge data science techniques, I solved real-world problems using supervised and unsupervised machine learning models and inferential statistics. Please see my Data Science projects below
Appointed to the Henry and Martha Zaytoun and Family Chair by world-renowned music director, Grant Llewellyn in 2019, I served as associate principal trumpet of the North Carolina Symphony. I led the brass section in weekly performances at Meymandi Concert Hall performing works by Beethoven, Brahms, and Mozart regularly. Prior to the appointment, I served as fourth / utility trumpet.
Connecting professional musicians with music lovers, I facilitate performances and execute contracts on behalf of both parties. I work with local and national performing arts organizations to produce some of the finest live music performances in central North Carolina.
I curated curriculum and presented weekly lectures on advanced performance topics and music theory. I advised students on career development in the performing arts.
Apart from being a data scientist, I spend most of my time with my family. I have two young children and an amazing/supportive spouse. I love powerlifting and being outside. My family spends a large amount of time in the Appalachian Mountains where we enjoy hiking and camping.
When not with my family, I spend a large amount of my free time exploring the latest advancements in the data science world, playing new video games and practcing trumpet