• Prototyped, evaluated, and productionized machine learning models at a global user scale; Examples include media mix models, time series forecasting, budget optimization, churn prediction, promo targeting, uplift modeling, and causal inference.
• Translated abstract business goals and objectives into concrete quantifiable metrics, performed metric evaluations, and designed A/B tests to measure metric changes.
• Designed and implemented a Python package to translate business logic into a constrained optimization problem, enabling 10+ data scientists to solve problems across products and regions.
• Led a bottoms-up transformation for a primary deliverable by providing a proposed architecture design and proof-of-concept which reduced reporting cadence from quarters to weeks. The new approach decreased data scientist hours on this task by at least 50%.
• Core developer and contributor for Orbit, an open-source Python package for Bayesian modeling and inference available now on PyPI and GitHub.

• Assist in the instruction of Statistical Methods for Discrete Response, Time Series, and Panel Data (W271)

• Graduated January 2019• Intensive courses in statistics, machine learning, and data storage and retrieval• For additional details see list of courses and projects

• Extracted and transformed heterogeneous data sources to create integrated panel data containing up to a thousand variables; collaborated closely with statisticians to define necessary variables.• Validated incoming client data by using both ad-hoc and systematic methods to generate time-series visualization and summary statistics using Tableau, Excel, and R.• Developed a data pipeline using Python and Hive, transforming terabytes of marketing data feeds on an automated monthly cadence.• Eliminated wasteful manual work by building a data ETL framework which led to a 30% increase in efficiency; prototyped a proof of concept and wrote technical requirements for engineers to implement.• Leveraged my breadth of knowledge to bridge the gap between data scientists and account managers to create a new product used to measure TV ad performance.• Generated search terms for Google Trends; developed a generalized approach for normalizing search volumes.

• Developed SAS programs to create analyzable datasets from data types with high variety; focused on producing both repeatable and ad-hoc processing.• Performed hypothesis testing to evaluate the disparity in employee wages between independent groups.• Conducted field experiments resulting in minute-by-minute data points of various laborers, which were used to analyze the rate of labor law compliance.• Produced economic damage estimation models using Excel, which enabled non-technical users to perform interactive scenario analyses.• Generated tables and visualization for legal filings.

• Cooperated with 26 other Orientation Advisors to represent USC for an incoming class that comprised of 4,500 freshmen and transfer students.• Led a student life discussion with a group of approximately 20 students for each orientation session.• Discussed information about the school and answered any questions for up to 700 students and guests.

• Founded a new chapter with 25 members and expanded to a 100 man house over a course of 3 semesters.• Budgeted and managed annual revenues that increased from $50,000 to $500,000.• Prepared the budget on a bi-weekly basis to all members of the fraternity.• Devoted 15 hours a week to volunteer and participate in executive board and alumni meetings.

• Created a database of all archived research using Microsoft Access and demonstrated the benefits of its use to the Project Manager.• Summarized the comprehensive economic development strategies for current and past clients.• Investigated various sources regarding data centers to analyze market trends.• Converted quantitative data into graphs, charts, and maps using Microsoft Excel and GIS software.