Xiaoyu Yang Email & Phone Number

Seattle, Wa, Us

A/B testing and observational studies to understand the incremental impact of Amazon Associates and Influencer Program

Woonsocket, Ri, Us

• Led the launch of a pioneering care management lite program to 200K Medicare plan users, driving 35% user behavior change and 10K incremental compliances in health screenings• Standardized the experimentation framework across 5 data science teams with 40+ concurrent digital campaigns, improving the accuracy of program-level impact by 30%• Leveraged Bayesian Structural Time Series model using Python to quantify the impact of digital campaigns without control groups, increasing team-level business impact by 10%• Collaborated with 3+ cross-functional agile teams and managed 5+ work streams with Rally / JIRA in a fast-paced experimentation environment• Led the presentation of monthly governance meetings to communicate insights and businessstrategies with technical and non-technical stakeholders in formal leadership updates• Hired, managed, and led a high-performing team of 3 data scientists that has successfully delivered 3 high-profile projects within 6 months

Woonsocket, Ri, Us

• Led the launch of a site-of-care digital campaign series to 800K commercial insurance plan users, driving 5.5-10% of user behavior change and $5-7.5M incremental medical cost savings• Designed 10+ A/B/multivariate tests and leveraged advanced techniques to improve the robustness of the experiments by 30-50%• Performed statistical tests and regression analysis to understand the incremental impact, facilitating the decisions to launch future iterations that drove 2-5% incremental user behavior change• Developed and deployed a multi-class neural network model using keras to predict plan users’ likelihood of needing minor surgeries in the next 6 months with an AUC of 87%• Mentored 5 junior colleagues and dedicated 1+ hours each week to share technical best practices, improving team's technical skills and boosting productivity by 15%

New York, Ny, Us

• Developed and deployed an xgboost classifier using R and a multi-server queueing model using Python to predict underwriting workload and facilitate staffing decisions, reducing the overall pricing turnaround time by 10%• Automated underwriter performance scorecard and operational bottleneck reports in Tableau to facilitate operation strategies, reducing the time spent on reporting by 90%• Presented as a representative of the data science team in a monthly strategic meeting to technical and non-technical stakeholders across the company

New York, Ny, Us

• Developed and deployed a logistic regression model and a K-Means clustering model in R to target customers for a cross-selling digital campaign of auto and home insurance products, driving 15% incremental conversion rate• Identified defective metrics and automated the QA process for an executive business intelligence project in Tableau, reducing the time spent on manual QA by 90%• Collaborated with engineering, UI design, and marketing to track projects and iterations using JIRA and followed an agile process

Master Of Science - Ms, Operations Research

Bachelor Of Arts - Ba, Economics