This dashboard is a high-performance, in-browser analytics engine processing over 25 million user reviews. To construct a complete analytical picture, the raw dataset was enriched via a custom Python data pipeline that fetched real-time metadata—including Directors, Studios, and Casts—directly from the TMDB REST API before compiling the results into highly compressed Parquet files.
The front-end leverages DuckDB-Wasm for lightning-fast, client-side SQL querying and D3.js for reactive visualization. The application utilizes omni-directional cross-filtering: brushing over chart ranges or typing into the cascading text filters instantly triggers Common Table Expressions (CTEs) that re-query and synchronize the entire application state without relying on a remote server.