warehouse-optimization

Data warehouse optimization applies clustering, partitioning, and materialization strategies to reduce query latency and compute costs. It profiles expensive queries using baseline metrics to identify bottlenecks across staging, marts, and core dbt models.

Yes, this optimization skill provides cross-platform guidance tailored to Snowflake, BigQuery, Redshift, and Databricks. It adjusts clustering, partitioning, and warehouse sizing recommendations based on your specific data warehouse platform.

The best way to tune slow dbt dashboard queries is to first run baseline tooling to profile expensive queries, then apply clustering, partitioning, and materialization changes to the underlying models, and finally validate improvements using the provided metrics tools.

You profile expensive dbt queries by running the included cli tools, cost-estimate.js and model-stats.js, which read context from your data stack. These tools generate baseline metrics that highlight performance bottlenecks and cost pressure.

You apply clustering and partitioning based on query patterns identified during baseline profiling. The skill analyzes your model stats and query performance to recommend specific clustering keys and partitioning strategies for staging, marts, and core models.