projects

Stamford COVID-19 Dashboard

A dashboard for Stamford, CT that displays case, death, test, vaccination, and hospital capacity data broken up by demographic descriptors (age, race, gender). This proved to be helpful in targeting hyperlocal campaigns for vaccinations and COVID-19 precautions.

Flask-Base

A Flask application template with the boilerplate code already done for you. Check out the article "How we got our 2-year-old repo trending on GitHub in just 48 hours" published in FreeCodeCamp. I'm currently the active maintainer for the open-source project.

SpineTK

Comprehensive toolkit for analyzing spine imaging studies. Detects and diagnoses vertebral fractures using neural networks (mask-rcnn w/keypoint rcnn and segmentation heads). Active research project in progress in Perelman School of Medicine Dept of Radiology.

Anti-Wage Theft Application

This application was developed for Community Legal Services to find new evidence for wage theft trials in Philadelphia. The website parses google location history locally and produces a virtual timesheet based on that data for use in the legal discovery process. Writeup at "How we built an app that uses worker location history to combat wage theft" for FreeCodeCamp.

XeniaPenn

XeniaPenn is a website I made to connect UPenn students (recently displaced by the pandemic from their dorms) to local students living off campus (so they could find housing and storage space temporarily). Coded in a few hours, it eventually helped hundreds of students find a temporarily place to live until they could arrange for travel off campus.

Auto Alt Text

I built this chrome extension to help visually impaired individuals get image descriptions when none are provided. Built on top of a modified version of im2txt and deployed as an API using AWS Lambda. Writeup at "Making the Web More Accessible With AI" for HackerNoon.

LeARn

Program for making collaborative AR experiences. Top 10 at PennApps XIII and Best VR/AR Hack. LeARn allows users to interact with physics simulations, plot 3D graphs, view MRIs, and watch the molecular structure of a chemical compound float before their very eyes. The user can move and manipulate these projections via an online client, affecting factors like scale, rotation, and position. Most importantly, when, say, a teacher makes these changes, all other users viewing the same object will also see them take effect.

MoleML

App for detecting Melanoma and Basal Cell Carcinoma using OpenCV/ML. Made at PennApps XII, Won Best Health Hack. User can take image with cell phone. Image is then sent to a server and processed by a OpenCV algorithms, Random Forest Classifier trained on a public, HIPAA-compliant data set containing over a thousand images of potentially cancerous moles. The algorithm returns its prediction and confidence, calculated by analyzing up to ten parameters of both the mole and the user.