• Analyzed millions of customers’ seasonal purchase behavior for festivals using Apache Spark and Seaborn• Worked on feature engineering and data cleaning including categorical encoding and sequence creation• Developed a Transformer Neural Network Model in TensorFlow to sequentially recommend top-k items• Performed hyper-parameter grid-search by deploying model on a Google Dataproc cluster and achieved an AUC of0.88, Mean Reciprocal Rank of 0.56, and Normalized Discounted Cumulative Gain of 0.8• Evaluated user, item, and time embedding quality by analyzing k-closest embedding by Euclidean distance

• Visualized diabetes severity by zip code using GeoPandas and Folium• Tested for statistically significant differences in HbA1c among demographic groups using ANOVA and Welch’st-test with Bonferonni Correction• Modeled HbA1c with LASSO and OLS regression models achieving an R2 of 0.77• Authored weekly written reports and presented insights to SDRI researchersProjects

• Handled and supported a wide variety of tax compliance processes• Prepared and filed tax returns for international clients • Worked with audit teams to determine whether tax accounts are appropriate for financial statements• Researched tax code and wrote memos for clients

• Built web-scraping programs using python and BeautifulSoup4 to crawl through thousands of webpages of state legislative data• Wrangled data such as bill id, sponsors, description, votes, and current status on every bill from 2000 – 2020 with Pandas