• Interpreted large-scale Medicaid claims and behavioral health datasets, applying statistical and analytic methods to identify utilization patterns and support state compliance decision-making across Wyoming’s statewide BHMS platform.
• Used Python and SQL to develop analyses, automated reporting pipelines, and data products supporting multi-year utilization analysis and provider comparisons for state health program managers.
• Translated ambiguous state agency reporting requirements into concrete, measurable data deliverables; communicated analytical findings and system capabilities to non-technical program managers and state stakeholders through reports and presentations.
• Designed and implemented scalable SQL Server schemas and data models used consistently across multiple behavioral health programs to support high-volume transactional and analytical workloads.
• Built HIPAA-compliant Medicaid claims ingestion pipeline in C#, normalizing raw state data into auditable datasets supporting compliance reporting.
• Authored complex SQL queries, views, and stored procedures powering downstream analytics; optimized performance through indexing and refactoring for large datasets.

Relevant coursework: Probability & Statistics, Machine Learning, Deep Learning, Data Visualization, Data Science Tools, Data Science Capstone

Relevant coursework: Regression Analysis, Categorical Data Analysis, Design & Analysis of Experiments, Applied Multivariate Analysis, Fundamentals of Sampling

• Developed analytical SQL views and Python scripts enabling utilization analysis, provider comparisons, and multi-year trend reporting for state program managers.
• Integrated Medicaid claims data using a custom C# ingestion service, linking external records with internal systems for longitudinal analysis and audit support.
• Implemented secure data handling, audit logging, and traceability to meet HIPAA and state compliance requirements.

• Built and evaluated supervised ML models in Python using TensorFlow, Scikit-Learn, and XGBoost, including text classification for political bias/sentiment analysis and predictive models on global public health datasets (World Bank, GTD).
• Applied NLP techniques including tokenization, embeddings, and sequence modeling; visualized model outputs and evaluation metrics using matplotlib, seaborn, and Plotly to communicate findings clearly.
• Designed reproducible end-to-end ML workflows - data ingestion, preprocessing, feature engineering, cross-validation, hyperparameter tuning, and performance monitoring - using modular Python code and version control.
• Produced analytical reports and visualizations in Power BI and Python to communicate results to technical and non-technical audiences.