Resources

BigQuery - SQL

BigQuery served as the foundation of my data pipeline and handled all of the heavy lifting required for this project. I used it as the primary data warehouse to process millions of rows from Google Trends’ Top 25 weekly search terms across all 210 DMAs. BigQuery was ideal for large-scale operations such as filtering four years of historical data, joining multiple demographic datasets, and extracting DMA rankings. Its distributed SQL engine allowed me to clean, normalize, and structure the data efficiently before exporting only the essential, analysis-ready tables.

Tableau Public - Desktop

Tableau served as the front-end experience for this project, bringing together all DMA-, state-, and national-level datasets into a single interactive dashboard. To support complex navigation across levels and ensure that interactions behaved correctly, I designed a structured Tableau data model with clean relationships, custom filtering logic, and parameter-driven controls. The result is a fully dynamic dashboard where every selection, whether a term, DMA, state, or category, updates all relevant charts simultaneously and without conflict.

VS Code- Python

After BigQuery handled the large-volume preprocessing, I moved into VS Code for the more customized and iterative parts of the pipeline. Python allowed me to roll the DMA-level data up to the state and national levels using refined population-weighting logic that would have been cumbersome, unreadable, or impractical in SQL. VS Code also enabled me to integrate the OpenAI API to automatically categorize all search terms into category labels. This stage let me apply flexible logic, run custom functions, and produce the final cleaned datasets that powered the Tableau dashboard.