My Programming Portfolio
Welcome to my programming portfolio! I specialized in computer vision and deep learning while getting my Master of Science at Columbia University.
I’m a backend software engineer at Dow Jones, working with GraphQL, Node.js, AWS, and TypeScript.
Fingertip Finder Featured Project
Fingertip Finder is a two-layer network that uses the input of an RGB image of a person pointing at a camera and accurately finds the coordinates of the hand via pixel segmentation. Fingertip Finder can then take that hand and use coordinate regression to find the fingertip to the nearest pixel. This approach can be run in real-time on a computer and has a variety of different use cases.
Bergle Featured Project
Bergle is a fully functional search engine for the Muhlenberg.edu website. This project was my final project for artificial intelligence. The techniques and code used for this project can be easily reimplemented to work on multiple websites on a far larger scale.
The LEGO Rating Predictor Featured Project
The LEGO Rating Predictor uses supervised data scraped directly off LEGO’s website to predict different LEGO set’s star ratings. This dataset consists of price, piece count, brick colors, themes, set names, and more. Using this information and the k nearest neighbor machine learning classifier, we were able to predict the star rating of different LEGO sets to the nearest tenth with a 97.3% testing accuracy.
