o Developed a Web Scraper Pipeline to collect SEC Filings from EDGAR website. Built an Extract, Transform, and Load (ETL) Pipeline to load structured and unstructured text data into SQL database, improved the efficiency of data sharing and data searching.o Built a text classification model on SEC Filings to detect data breach with Logistic Regression and NLP. Engineered natural language features using NLTK and spaCy. Vectorized text data using TFIDF, and LSI/LSA/SVD with Scikit-learn. The Logistic Regression model achieved AUCROC of 93%.o Built a regression model to predict the timeliness of data breach disclosure with Linear Regression model. Selected relevant corporate governance features and data breach features. Statistically examined causality between timeliness of data breach disclosure and a certain corporate governance feature, Audit Committee Cybersecurity Risk Oversight. The result suggests that Audit Committee expedites data breach disclosure, and provides insights to regulators and companies about the role of Audit Committee in Cybersecurity Risk Oversight. o Managed a team of four data engineers to collect non-GAAP Financial Measures from Earnings Announcements.o Developed a Linear Regression model to predict comparability of non-GAAP Financial Measure using SAS. Identified relevant determinants and provided explanation for each determinants.o Parsed analysts’ names and titles from analyst reports using regular expression.

Taught Financial Accounting, managed a class of 90 students, and held regular office hours. Achieved a teaching evaluation of 4.2/5.

o Developed an Interactive Pricing Engine for real estate agents. Cleaned and processed 3GB rental house dataset and engineered new features.o Built and optimized rental income prediction models using tree-based models (Random Forest, XGBoost). o Wrapped the model with interactive interface as an API using Flask, allowing clients to adjust the model, and deployed it to Google Cloud Platform (GCP).