AI & Health Data Scientist
Maggie 'Xiaomeng' Xu
I’m passionate about exploring the potential of AI to make everyday life smarter, healthier, and more human-centered. Through data science, statistical modeling, and creativity, I aim to turn complex data into meaningful change for people and communities.
About me
Welcome to my data-driven journey. I’m fascinated by how artificial intelligence and statistics can reveal hidden patterns in the world around us, from the pulse of human health to the rhythm of everyday life.
To me, every dataset is a story waiting to be understood. Each model, each graph, is a step toward connecting insight with impact, turning numbers into knowledge that helps people live better.
If you’re curious about how AI and data science can make a real difference, let’s explore together.
Work experience
Assistant Statistician
Bauer Research Group, University of Michigan, Ann arbor, MI
May 2025 - Present
Collaborated with PI to define research objectives and translate scientific questions into appropriate statistical analyses.
Designed and executed statistical analyses using logistic regression, ANOVA, andcorrelation methods to assess health outcomes and demographic associations.
Produced statistical summaries and figures (TLFs) in SAS and ensured reproducibility of results across analyses.
Delivered statistical tables and figures supporting a manuscript on self-regulation and family feeding behaviors.
Data Analyst
Center for Academic Innovation, Ann arbor, MI
Sep 2024 - Aug 2025
Analyzed large-scale educational datasets using R and SQL to identify equity gaps in student performance across demographic groups, generating actionable insights for institutional policy.
Built interactive R dashboards to visualize performance trends, improving data accessibility and evidence-based decision-making for non-technical stakeholders.
Automated SQL and R-based ETL pipelines for 1M+ student records, cutting dashboard refresh time by 45% and improving workflow scalability.
Utilized Git-based version control and collaborative workflows to maintain analytical integrity and support cross-functional research teams.
Education
Master of Science in Health Data Science
University of Michigan's School of Public Health
2023 - 2025
Bachelor of Science in Food Science
Southwest University, Chongqing, China
2018 - 2022
Bachelor of Science in Finance
Southwest University, Chongqing, China
2019 - 2021
Skills
Core technical and analytical skills that empower my research and data-driven work.
Statistical Modeling
Regression, ANOVA, survival analysis, mixed models, and hypothesis testing.
Machine Learning
Supervised and unsupervised learning, feature selection, model evaluation, and explainability.
Deep Learning
Neural networks for image and tabular data; transfer learning and model tuning.
Programming
R, Python, SQL, and SAS for data cleaning, wrangling, automation, and reproducible research workflows.
Data Visualization
Creating clear, interpretable dashboards and figures using ggplot2, Plotly, Power BI, and RShiny.
Data Engineering
Experience with data pipelines, EHR and wearable data integration, and relational database management.
Communication and Collaboration
Translating analytical results into actionable insights; effective teamwork across multidisciplinary teams.
Attention to detail
Diligent in maintaining precision and quality in all analytic and research work.
Adaptability
Quick to learn new methods, technologies, and analytical frameworks in evolving data environments.
Bridging Data and Humanity
Projects that combine analytics, design, and empathy to turn complex health data into stories that inspire understanding and action.
Project 1:
Personalized Cardiovascular Risk Scoring System
Project Summary:
Developed a personalized cardiovascular risk scoring system that integrates electronic health record (EHR) data from Michigan Medicine to create an interactive, interpretable health analytics experience. Using machine-learning models such as XGBoost and SHAP, the project transforms complex medical data into meaningful insights, helping individuals understand how their daily choices and physiological factors influence heart health.
Role:
- Developed the end-to-end system, from EHR data preprocessing and feature engineering to model optimization and front-end implementation.
- Built the XGBoost-based predictive model with SHAP interpretability.
- Designed an intuitive dashboard that visualizes personalized cardiovascular risk, contributing factors, and actionable insights for preventive decision-making.
AI & Statistical Consulting
I offer free 30-minute consultations for researchers, students, and organizations exploring AI and data analytics solutions. Together, we can turn data into insight, and insight into impact.
Learn More & Book Free Consultation