Experience

• Research for Chan Zuckerberg Initiative (CZI): inferred 30-40% of U.S. adults with criminal records, ~25M people, are eligible for criminal record expungement. Insights helped CZI lobby regulators for state and federal legislation of automatic record clearance solutions.
• Gathered continual insights on anonymous criminal record dataset of ~60K records to relay to CZI to ensure continued funding of project.
• Tesla Motors: created Python library for drive simulation hardware to programmatically control car’s 12V battery; tests used to validate various aspects of Model 3’s design.
• Designed and taught curriculum for 11-week master’s business analytics course in Python at Santa Clara University for 30 students; received positive reviews from students.
• Created Random Forest classifier predicting trial rate conversion for personal trainer app with ROC AUC score of 0.88; model assisted team to interact with leads more efficiently.
• Discovered drop-offs in user onboarding flow for trainer signups on app; suggested recommendations that were adopted by developers to redesign flow.

• Designed and implemented lead scoring model using engagement in social network app that estimated student college enrollment likelihood; guided Director of Admissions on outreach.
• Implemented XGBoost classifier that predicted admitted college students’ likelihood to enroll.
• Analyzed trends in customer data and produced 30+ data visualizations using Python and Plotly that were featured in customer-facing product.
• Identified bottlenecks in email deliverability product through SQL queries and development of internal Chartio dashboards; insights used by support for diagnosis and helped engineers identify bugs and reduce time to send.
• Created metrics and visualizations to highlight product value for renewal sales decks; helped close multiple $15K+ annual contracts.

• Provided evidence for engineers to make hardware and software changes before production by identifying outliers and issues in riding telemetry.
• Created KPIs for engineering teams evaluating product performance from test riders highlighting engineering efforts towards hardware launch goals.