MS Business Analytics & AI @ UT Dallas · Dallas, Texas
I build forecasting, reporting, and decision-support systems for operations, supply chain, and business performance. My focus is on work that connects data to real decisions — not just technical outputs.
- Forecasting & ML — demand forecasting, regression, classification, time-series models
- SQL & Data Architecture — schema design, analytical queries, CTEs, window functions
- BI & Reporting — Power BI dashboards, DAX, KPI frameworks, automated reporting
- Operations Analytics — supply chain analysis, inventory planning, cost variance, margin analysis
An end-to-end analytics system combining forecasting, risk classification, throughput prediction, SQL architecture, and SHAP explainability — deployed as a live Streamlit app.
- R² = 0.874 throughput model
- $2,286 estimated cost per disruption hour
- 53,000+ supply chain records analyzed
- 6 statistical validation tests in R
An end-to-end AI agent that automates the monthly FP&A workflow — ingesting actuals vs. budget data, running multi-period variance analysis in SQL, detecting anomalies with adaptive machine learning, and generating board-ready management commentary using an LLM. Designed to replicate real FP&A workflows used in finance teams, and built from hands-on experience doing this work manually across 20,000+ financial records.
- Adaptive ML-based anomaly detection with stability-aware logic to avoid false positives on clean datasets
- SQL variance engine built with CTEs, window functions, and FULL OUTER JOIN-style logic to preserve budget-only and actual-only records
- Board-ready management commentary generated through an LLM pipeline using the Groq API
- Full analysis completed in under 2 minutes, reducing a manual 2–5 day month-end workflow to a repeatable automated process
