• Managed global analytics partner, Analytic Edge. This included designing and deploying a MMM to optimize our ~$30 million FY25 marketing budget plans, forecasting an improvement of $1.25 million in sales despite shrinking our budget by $2 million.• Gathered intel from marketing, ecommerce, analytics, and third-party teams to refine the sales MMM by incorporating said intel as Bayesian priors, resulting in forecasting the next two-month sales with only 3.5% error.

• Led redesign of a modern, weighted multi-touch attribution (MTA) model to optimize our ~$25 million/year digital spends and improve efficiency by more than 10% in a declining budget and sales environment.• Collaborated with Chief Marketing Officer to plan fiscal year budgets using MTA model, increasing ROI efficiency by 10% in FY23 and then 1.4% in FY24.• Took charge in mentoring 7 interns, training them to communicate clearly and focus on high value projects (such as pricing… Show more • Led redesign of a modern, weighted multi-touch attribution (MTA) model to optimize our ~$25 million/year digital spends and improve efficiency by more than 10% in a declining budget and sales environment.• Collaborated with Chief Marketing Officer to plan fiscal year budgets using MTA model, increasing ROI efficiency by 10% in FY23 and then 1.4% in FY24.• Took charge in mentoring 7 interns, training them to communicate clearly and focus on high value projects (such as pricing elasticity and product recommendation systems). Show less

• Designed, tested, and coded a propensity score matching algorithm to help our broadcast team measure loyalty campaign performance, resulting in doubling the spend from $5 to $10 million.• Using the propensity score matching algorithm, guided our Chief Analytics Officer to perform causal inference on website waitlist changes, customer retention, and other customer interventions, improving our long-term customer retention strategy. • Selected to work directly with co-founder to… Show more • Designed, tested, and coded a propensity score matching algorithm to help our broadcast team measure loyalty campaign performance, resulting in doubling the spend from $5 to $10 million.• Using the propensity score matching algorithm, guided our Chief Analytics Officer to perform causal inference on website waitlist changes, customer retention, and other customer interventions, improving our long-term customer retention strategy. • Selected to work directly with co-founder to generate customer analytics reports resulting in 4 new customer journey campaigns Show less

• Saved $18,000 per employee per year by founding and building a segmentation customer analytics dashboard• Led research and reallocation of dormant assets worth ~$100,000 into a multi-GPU cloud server • Tasked by the CEO to perform behavioral customer segmentation analyses on high-value customers, applying machine learning models to answer executive needs

• Improved coursework for Linux classes with new focus on VI text editor and regular expressions• Learned SQL to be the head SQL teaching assistant for next term• Supported professor and head TA in creation of Spark/Hadoop Big Data Ecosystem Class

• Updated the tutorials and moving course work from old website to new website• Managed other teaching assistants’ shifts based on students’ needs• Hosted teaching sessions (Tidyverse) 10 hours per week to read and debug code

• Predicted long-term care insurance claims from a huge dataset (8M rows) and learned that generalized linear models best fit the data• Worked on building a k-nearest neighbors machine learning algorithm from scratch to predict claims (~20 hours)• Collaborated with other assistants and professors weekly

• Used Neural Network software to find early adopters and investors for the company• Met weekly with the CEO to work on visualization and regression projects• Supported engineering teams by learning how to do simple web scraping in Python

• Delegated and managed over 300 hours of research to commercialize a new technology and present a pitch deck to investors and venture capitalists • Created and distributed a Qualtrics survey for market research on customer values with over 80 responses • Helped find and analyze data for customer yearly value, quality vs. price product comparisons, and cost structure

• Refined evaluation system by developing new language assessment• Increased students’ Chinese vocabulary by 60% in three weeks through implementing new curriculum• Promoted greater accountability by automating language self-tracking through excel