Data Scientist & Data Analyst · Lakewood, WA · Open to remote, hybrid and on-site roles
I turn messy business data into reports, models and automations that people use every day. Most recently I built the automated daily sales reporting for a data analytics team: scheduled Excel reports, a WhatsApp report bot, and a web tool that lets staff fix data-quality issues before the next run.
| Project | What it shows |
|---|---|
| Meraki Sales Reporting & WhatsApp Automation | Production n8n + Python pipeline: daily Excel reporting, WhatsApp/email delivery, a data-quality resolver web app, an AI tracker chatbot |
| Seattle 911 Data Pipeline · live | Daily automated ETL on GitHub Actions: open-data API → DuckDB → SQL → 10 data-quality checks → self-publishing dashboard |
| Olist E-Commerce SQL Analysis | 15 business questions in SQL on 96,000 orders: cohort retention, delivery vs. reviews, seller scorecard, RFM segments |
| Washington EV Adoption Dashboard · live | Tableau dashboard of 299,353 EVs: per-capita county map (LOD expression), top makes, growth by model year |
| Customer Churn Prediction · live app | Churn model (ROC-AUC 0.84) with a profit-based contact threshold, deployed as a Streamlit app |
| Customer Segmentation (RFM + k-Means) | 4,245 retail customers segmented with SQL and k-means, backtested a year ahead (94% of Champions bought again), with a marketing action per segment |
| Cookie Cats A/B Test | Experiment analysis on 90,189 players: sample-ratio check, z-tests, bootstrap, Bayesian and power analysis, unit-tested stats code, and a ship / don't-ship call |
| 10-Year Coronary Heart Disease Risk | End-to-end ML on the Framingham study: leakage-safe pipeline, PR-AUC model selection, threshold tuning, interpretation |
| Superstore Sales Forecast | Monthly sales time series and a 12-month ARIMA forecast, exported to Excel |
| Ames Housing Prices | Regression on 2,930 home sales |
| COVID-19 Vaccination Progress | Interactive Plotly trends across five countries |
- Analysis & ML: Python (pandas, NumPy, SciPy, scikit-learn, statsmodels, imbalanced-learn), Jupyter
- Statistics & experimentation: A/B testing, hypothesis tests, confidence intervals, bootstrap, Bayesian A/B, power analysis
- Deployment: Streamlit apps on Streamlit Community Cloud
- Data & reporting: SQL (DuckDB, PostgreSQL), Excel, Tableau (public profile), Power BI
- Visualization: Matplotlib, Seaborn, Plotly
- Automation & data engineering: n8n, GitHub Actions (scheduled pipelines, CI), REST APIs and webhooks, data-quality checks, WhatsApp (WAHA), SMTP, ClickUp