Take a look at my resume:
Built a site to fulfill a business need within the Teamfight Tactics community. Leveraging SEO and organic marketing, was able to go from 0 to 60,000 visits within 2 months. Currently hosting legacy version of site on Netlify.
Air pollution is an almost ubiquitous problem with significant effects on human health, so we sought to better characterize the behavior of quantifiable air pollution. We investigated the effect of weather on the amount of different fine particulate pollutants in the atmosphere. We used the body of air quality data from OpenAQ, cleaned it, and paired it with historical weather data. We trained different machine learning classifiers to ultimately predict air quality changes with over 70% accuracy. We used Python for cleaning the data and for the machine learning.
dHOPE engages users by keeping track of their DPP (Drain Protection Points) in order to incentivize community members to take initiative in taking care of their local drains. Because these regions often lack the infrastructure to maintain up to date databases for municipal drainage systems, we used a crowdsourcing solution to identify the drains. In order to engage the community with the idea of protecting their local drains, we designed dHOPE to be a fun, interactive experience. Members of the community are awarded points for identifying, confirming, and adopting drains, and a local leaderboard tracks those who contribute the most to the community. In order to provide additional functionality, we also added additional features of a color-coded system and a weather panel to warn of imminent rainfall. dHOPE is built using meteor.js in conjunction with Javascript and HTML. We integrated multiple functionalities, with a MongoDB backend database used to store user and drain information. We implemented the Google Maps API in order to visualize our content and provide interactivity.
We won the Beginner Prize at the 2016 International Development Hackathon.
I was interested in examining so called 'hysteric' literature and seeing what characterized the novels and the characters as hysterics. Using Python and the Natural Language Processing Toolkit, I constructed ML classifiers on writing style and n-grams to examine what literary features could be used to classify a fictional "hysterical woman".