Compare Tableau live vs extract and filters

For a Tableau dashboard built on a fact table with 100 million rows, contrast Live and Extract connections. Cover refresh frequency, query pushdown, extract size, incremental refresh, row-level security, and the situations where each option is the better choice. Describe the order of operations and its impact on:

• Data source filters, context filters, dimension and measure filters, TOP/N, table calculations, and FIXED, INCLUDE, and EXCLUDE LOD expressions.
• Performance tuning: choosing context dimensions, applying extract filters, avoiding costly table calculations, and denormalizing versus using joins or relationships. Translate these concepts to their Amazon QuickSight counterparts (SPICE versus direct query, row-level security, filters), and identify one pitfall when moving a Tableau workbook to QuickSight. Overview: This question tests BI data architecture and large-scale dashboard performance skills, including extract versus live connections, refresh and incremental strategies, query pushdown versus in-tool processing, row-level security, filter order of operations, performance tuning, and platform migration mapping within the Analytics & Experimentation domain for a Data Scientist role. It is often used to evaluate trade-off reasoning across data freshness, storage, query performance, and security for a 100-million-row fact table, blending practical implementation and tuning decisions with conceptual understanding of filter ordering, calculations, and migration pitfalls when switching between BI platforms. This question is drawn from an Amazon Data Scientist interview.