Mileva Van Tuyl

Software Engineer, Data Scientist

Portfolio

AI Fairness 360

Data Science Intern, IBM

Trustworthy AI, machine learning, R package development

The AI Fairness 360 toolkit is an open-source library to detect and mitigate bias in machine learning models throughout the AI lifecycle.

More: Project Website, Open-source code, Documentation

The AI Fairness 360 package is used to reduce bias according to five fairness metrics.

Data for Diabetes

Data Scientist, Minerva Analysis

Machine learning, data analysis, full-stack development

Minerva Analysis equips type I diabetics with tools to regulate their blood glucose levels and improve their health outcomes.

More: Project Website

Saving Face

Project Co-Lead, MIT Media Lab

Machine learning, research, signal processing, mobile development

The Saving Face app alerts users when they touch their faces and provides a low-cost solution to reduce hand-to-face transmission of COVID-19.

Publications: IMWUT '21
Presentations: Harvard NCRC '21, Ruhlman Conference '21,
UMass Amherst Voices of Data Science '21
More: Project Website, GitHub code

Advised by Dr. Camilo Rojas, Fluid Interfaces

NYC Mobility Patterns

Data Scientist

Data engineering, data analysis, distributed computing

The NYC Mobility Patterns project provides key insights about factors that influence ridership on ridesharing platforms like Uber and Lyft.

More: GitHub code

Choropleths were used to analyze ridership on ridesharing apps throughout New York City.

COVID-19 Tracker

Developer

Data visualization, dashboarding, R Shiny

The COVID-19 Tracker is a dashboard that enables users to visualize the latest trends in COVID-19 cases and deaths across the United States.

More: Dashboard, GitHub code

The COVID-19 Tracker Dashboard displays statistics and visualizations on the total number of cases and deaths in the United States since Winter 2020.

News Nutrition Label

Undergraduate Research Assistant,
Wellesley College Cred Lab

Data science, research, web scraping

The search engine “nutrition label” (under development) is a browser extension that helps users distinguish credible from suspect media sources.

Presentations: Ruhlman Conference '19, Research Poster '19
Press: Wikipedia Edit-a-Thon

Advised by Dr. Eni Mustafaraj, Computer Science

The nutrition label supplements Google search results with credibility signals to help users discern the trustworthiness of the source.

Perfect Placement

Developer

Software development (Java)

The Perfect Placement software for college administrators implements a perfect matching algorithm to optimally place students in their preferred first-year writing course.

More: GitHub code

Perfect placement traverses a bipartite graph to optimally match students and courses.

Chemistry of Cosmic Ices

Undergraduate Research Assistant,
Wellesley College Physical Chemistry Lab

Research, data analysis

Analysis of mass spectrogram data demonstrates how prebiotic molecules, the building blocks of life, can be formed in interstellar ice clouds.

Publications: Royal Astronomical Society '19
Presentations: Northeast Chemistry Conference Poster '18

Advised by Dr. Chris Arumainayagam, Chemistry