Data Analyst | SQL, Python, Tableau, dbt, Airflow, PySpark | Turning messy, real world data into decisions leadership can act on.
I build end to end analytics workflows, not isolated scripts. From raw ingestion and pipeline orchestration through transformation, testing, and executive facing dashboards, each project below reflects the full analyst lifecycle a business actually runs on.
I recently completed my Master's in Computational Science and Data Science at UNC Charlotte. I'm a data analyst with a background spanning SQL and Python driven analysis, workflow orchestration, and stakeholder reporting. My projects are built the way production analytics work actually happens: messy source data, real infrastructure decisions, documented failures and fixes, and a final deliverable meant for someone outside the project to read and trust.
| Project | Description | Tools |
|---|---|---|
| Unified Commerce Data Pipeline | End to end e-commerce analytics pipeline: synthetic data with injected real world messiness, idempotent Postgres loads, dbt layered transformations, Airflow orchestrated DAG, PySpark trend analysis, and a live Tableau dashboard | Python, PostgreSQL, dbt, Airflow, PySpark, Tableau, Excel |
| Large-Scale Log Analysis & Anomaly Detection | Hybrid anomaly detection pipeline on 1.88M NASA HTTP server logs, combining Isolation Forest for point anomalies and ARIMA for temporal deviations, cutting false alert rate 39% versus threshold based alerting | Python, PySpark, Scikit-learn, Statsmodels, Matplotlib |
| Product Experimentation Analytics Engine | A/B testing framework combining frequentist and Bayesian approaches with power analysis and cohort segmentation to support product decision making | Python, Statistical Testing, Cohort Analysis |
| End-to-End Supply Chain Analysis | Supply chain analytics on a real world dataset using DuckDB for transformation, an automated PDF executive report, and Tableau visualization | DuckDB, Python, reportlab, Tableau |
| NovaMart Supply Chain Intelligence Platform | Star schema supply chain model with SQL window function analysis, Random Forest demand forecasting, and Power BI dashboards | SQLite, SQL, Random Forest, Power BI |
A full stack analytics pipeline built to mirror what a commerce analytics team actually runs in production.
- Synthetic e-commerce data with deliberately injected messiness: duplicate records, inconsistent date formats, orphaned foreign keys, late arriving records
- Idempotent Postgres loads via Python and psycopg2
- dbt staging, intermediate, and marts layers with automated tests
- Dockerized Airflow DAG orchestrating load, transform, test, and PySpark rolling revenue trend calculation
- Live Tableau Public dashboard and an Excel executive summary workbook
| Category | Tools |
|---|---|
| Languages | Python, SQL, R. |
| Databases | PostgreSQL, SQLite, DuckDB, MySQL |
| Transformation & Orchestration | dbt, Apache Airflow, PySpark. |
| Visualization & Reporting | Tableau, Power BI, Excel, reportlab |
| Analysis | A/B Testing, Statistical Modeling, Time Series, Anomaly Detection |