BigQuery Analytics

Build on Google BigQuery — partitioning, clustering, BI Engine, materialized views, BigFrames, Dataform, and slot management. Generates optimized SQL, governance configs, and cost-aware schemas.

This skill writes BigQuery SQL with proper partitioning and clustering, designs schemas to minimize scan cost, builds materialized views and BI Engine reservations for sub-second dashboards, uses BigFrames for pandas-like Python, writes Dataform/dbt models, configures slot reservations vs on-demand pricing, applies row-level and column-level security, and uses BigQuery ML and Gemini in BigQuery for in-warehouse ML and AI.

When to use

Use when migrating to BigQuery, optimizing expensive queries, setting up cost controls and reservations, modeling new datasets, or building ML and AI features directly in the warehouse.

Examples

Reduce a $400/day query

Partition pruning and clustering rewrite

Audit this BigQuery query that scans 8TB daily — recommend partition pruning predicates, clustering keys, and a materialized view to bring cost under $20/day without changing the consumer contract

BigQuery ML forecast

In-warehouse time-series model

Train an ARIMA_PLUS model in BigQuery ML on our daily revenue table, generate 30-day forecasts with confidence intervals, and expose as a view for Looker