Joy(Xinran) Tu Email & Phone Number

Houston, Tx, Us

Houston, Tx, Us

Houston, Tx, Us

San Diego, California, Us

1.Gathered and analyzed requirements for an AI-driven content curation and recommendation system by conducting competitive analysis, defining target user segments, and evaluating content types to optimize user engagement and platform effectiveness.2. Worked with cross-functional teams to design, implement, and measure AI-powered features, while exploring various pricing strategies, revenue models, and budget forecasts to maximize profitability and support long-term planning.3. Facilitated communication between technical and non-technical stakeholders, and leveraging data-driven insights to continuously improve the platform's AI-driven initiatives, ensuring alignment with project goals and requirements.

St. Louis, Mo, Us

Project: Marketing Campaign Optimization (Client in Online Food Delivery Industry)Analyzed marketing response data from the delivery app on their past year of email campaigns, compared offer response rates, and email campaign metrics, including open, click-through and conversion rates.Implemented a Decision Trees Model using Python to predict the likelihood to respond to wave 1 of incentive offers, enabling targeted marketing for wave 2 offers, which results in an estimated profit of $206K.Project: Advertising Injustices Analysis (Client in Fintech Industry)Analyzed potential injustice on advertising by banks and payday lenders to analyze the representation of different racial groups on their ads.Prepared the dataset for analysis by building data processing pipeline that uses AWS services (eg. Translations, Rekognition) to extract key metadata from various lending ads.Compared the percentages of different racial groups in lender’s ads vs. Annual County Resident Population Estimates, and conducted hypothesis testing on whether there is significant difference between them.Project: US Presidential Election ResultAnalyzed ​​election results by country from 2000-2020 and combined with external census and demographic information to examine voting patterns and trends over timeGenerated interactive visualizations in Tableau to visualize voter’s behaviours including vote share by party, margin of victory, swing state, voter’s demographic breakdown, voter turnout, etc.

• Pulled out 6k+ paid subscription data from 6 databases using MySQL to get order numbers of each product, show top 10 popular products from 6 Electronics categories and plotted the trends for next following 3 months, improved the efficiency of product selection for the next quarter by 50%• Took 5% of customers for sample RFM analysis in R, found the profitable cells and offered deals to all remaining customers who fell into those cells, increasing the subscription rate by 15% and saving 18% budget• Applied K-Means clustering model using Python to classified customers: used Elbow method to determine the K value, classified customers into K segments based on correlation measures and achieved 80% accuracy; Improved the efficiency of market campaigns for the new product and boosted sales by $60K• Performed Cobb-Douglas Model with 2GB data using R, provided an optimal market plan of labor and investment and reduced decision-making time for managers by 80%

Master'S Degree, Business Analytics

Bachelor'S Degree, Statistics