I’m a Stanford University graduate with a B.S. in Symbolic Systems (translation: Computer Science + Linguistics/Philosophy/Psychology with an AI concentration) and a minor in Economics.
My work and passion revolve around leveraging data to uncover insights that empower people.
I am currently a Machine Learning Scientist at Apple, previously at PayPal. There, I focus on training, evaluating and shipping models that improve Personalized Hey Siri. Simultaneously balancing state of the art methods and user privacy, I have been able to release models to more than 25 locales (and counting) on iPhone, HomePod and Apple TV. At PayPal, I worked on NLP and time series ML models (patent pending) that improved resource scaling efficiency, issue detection, and reliable troubleshooting of customer issues. I also worked across departments to convene a company-wide machine learning society in order to facilitate development and sharing of ML tools and best practices.
While in school, I led research on detecting workplace sexism using state-of-the-art Natural Language Processing (NLP) deep learning techniques on an original (non-Twitter) repository of workplace sexist statements. In the end, my machine learning algorithms were able to identify workplace sexist comments with an accuracy of over 88%, and the resulting paper I authored was accepted to the PAKDD 2020 Workshop on Learning Data Representation for Clustering.
I used the lessons of this research to help develop a prototype of a tool that detects gender bias in political reporting under Maneesh Agrawala, Krishna Bharat, and R.B. Brenner. The resulting paper was accepted for publication at The 2020 Computation + Journalism Symposium.
I continued my research with the Sustainability and Artificial Intelligence Lab under Marshall Burke, Stefano Ermon and David Lobell. My group’s project involved building novel methods to aid resource allocation in developing regions by using computer vision on crowdsourced images to predict wealth and health assets. The resulting paper got accepted to AAAI ’21.
At The Stanford Daily (our student-run newspaper), I founded the Data Team section and served as the paper’s first Data Team Director, expanding the section up to 25 data journalists. On top of writing and editing content for the @94305 data blog, I created and managed projects including the Stanford Open Data Portal, Modeling The Draw, the quarterly Stanford Community Survey, the yearly Stanford Senior Survey, and The Daily’s COVID-19 visualization coverage of the Bay Area. I also wrote about music!
In my free time, I love playing guitar, listening to music, reading FiveThirtyEight, keeping up with Stade Rennais FC, and using Oxford commas.