Western Governors University — M.S. Data Analytics, Data Science
Data Analyst · Data Scientist
Nearly a decade of multi-site operations leadership translating KPI, revenue, labor, and throughput data into reporting, decisions, and process improvement — now applied across a public portfolio of analytics, machine learning, forecasting, deep learning, databases, optimization, and MLOps work.
These cover the core story of the portfolio: end-to-end ML engineering, supervised classification on a real-world business problem, a published BI dashboard, and a hybrid analytics project that crosses supervised, unsupervised, and dimensionality reduction methods.
End-to-end ML engineering pipeline replicating a production MLOps workflow: versioned data with DVC, experiment tracking through MLflow, containerized deployment via FastAPI and Docker, and automated build and test through GitLab CI/CD. A live REST API serves real-time predictions from the Dockerized FastAPI endpoint.
Binary classification pipeline on 148,670 loan records with feature engineering, class imbalance handling via balanced class weighting, stratified k-fold cross-validation, and tuned Random Forest modeling. Outputs are framed as a business risk decision tool, not just a confusion matrix.
Interactive churn dashboard published to Tableau Public. Surfaces churn drivers by contract type, income, geography, and payment method, supported by two executive-facing KPIs and a color-blind-safe palette tuned for accessibility.
Three analytical methods applied to hospital readmission data: Random Forest classification, K-Means clustering for unsupervised patient risk segmentation, and PCA for dimensionality reduction and cluster visualization — demonstrating breadth across supervised, unsupervised, and dimensionality reduction techniques.
The broader portfolio shows range: independent regional market analysis, time series, relational and NoSQL database design, text modeling, computer vision, operations research, and applied statistical work on real datasets.
Regional housing-market screening framework comparing 105 Southeast metros on affordability, growth, and labor-market depth. Produces overall and balance rankings that separate low-cost markets from markets with stronger fundamentals.
ARIMA forecasting on 731 daily revenue observations with ADF stationarity testing, differencing, parameter selection, and holdout evaluation.
Normalized 3NF PostgreSQL design across five tables and 100,000 records with ETL staging, indexes, CTEs, and reusable analytical views — plus a parallel MongoDB implementation with aggregation pipelines and compound indexes.
Bidirectional LSTM sentiment classifier with sequence preprocessing, tokenization, padding, and evaluation on a labeled test set.
Convolutional Neural Network for multi-class classification of plant seedling species, with trained model artifact included for inference and inspection.
Five notebooks covering linear regression, logistic regression, PCA, hypothesis testing (ANOVA, Mann-Whitney U, Chi-Square) on health insurance data, and Apriori market basket analysis on 8,234 retail transactions surfacing high-lift association rules.
Linear programming formulation of a multi-node cargo transshipment network in PuLP, with capacity constraints and explicit verification of feasible flows.
Not an exposure list. Every tool below shows up in one or more of the ten public portfolio repositories, used on real data to produce real outputs.
Nearly a decade managing multi-site operations, KPI tracking, labor, throughput, and performance reporting. The portfolio work is built to support decisions — not just produce models — because that is what the day job has always required.
Xtreme Car Wash · Anniston, AL
Xtreme Car Wash · Anniston, AL
International Car Wash Group · Gadsden, AL
Western Governors University
Coursework: Machine Learning, Statistical Analysis, Data Mining, MLOps, Deep Learning, Optimization, Database Management.
Jacksonville State University
Summa Cum Laude · Special Honors in Management.
Gadsden State Community College