Aug 2022 – Jan 2025 · 2 yrs 5 mos · Remote

• ● Developed a Pytorch computer vision model that uses DINOv2 embeddings to predict the “fanciness” of a
• home based on interior real estate images; the model helps insurance carriers adjust premiums
• ● Fine-tuned an EfficientNet computer vision model to detect the presence of chimneys from aerial imagery
• at an accuracy of 88% to help insurance carriers adjust premiums
• ● Built a KNN model and data engineering pipeline in python that imputes missing values for 29 house
• features on 72 million single-family homes in the US; the imputed values are a core part of Cape’s prefill
• product that auto-fills house information for home insurance application

Jan 2021 – Jun 2021 · 5 mos

• ● Led a team of 4 data scientists and formulated the quarterly OKRs to ensure that projects were reasonable
• scoped and our team better supported the operations, finance, and customer-facing analytics teams
• ● Created a time series model for the operations team that forecasted the amount of human labeling work
• for the next few weeks, which reduced the amount of paid labelers who were idle by 62%
• ● Collaborated with the customer facing analytics and engineering teams to develop and release a consumer
• lender analytics product, which has a suite of 30 cash flow measures for loan underwriting
• ● Recruited 2 data science interns, helped them complete successful data projects, and convinced leadership
• to convert them to full-time employees

Jul 2019 – Jan 2021 · 1 yr 6 mos

• Recommended opportunities to automate document verification
• ● Created a deep learning NLP model that predicted the industry classification of small businesses with 60% accuracy using Python, tensorflow, numpy, pandas, and Fasttext; the model was deployed and supports a new credit risk product to help small businesses make better lending decisions
• ● Built and maintained ETL pipelines that processes 100,000 financial documents every day using Python, SQL, AWS, Postgresql, and Redshift
• ● Developed the business logic for measuring company-wide north star metrics including document accuracy, document volume, and turn-around time; the new reports help operations to process client documents on time and accurately

Mar 2019 – Aug 2019 · 5 mos · Greater New York City Area

• ● Mentored 50 students on python, statistics, data visualization, and machine learning
• ● Explained complex data science ideas to students with enthusiasm and empathy

Sep 2018 – Jan 2019 · 4 mos

• ● Built a web app that helps new yoga instructors find good locations to open new yoga studios
• ● Designed a logistic regression model that predicts which zip codes in New York state will open a yoga studio from demographics data with a recall rate of 74% using Python and scikit-learn

Jul 2013 – Jan 2019 · 5 yrs 6 mos · Houston, Texas Area

• ● Discovered neurons in the rodent brain that fire according to the 3D head orientation of the animal. These neurons are theorized to form a neural compass that is critical to navigating in 3D space.
• ● Published dissertation research in Nature Communications, a high-impact open access journal