Databricks SQL Autotuner

Analyze, rewrite, validate, and benchmark SQL queries against a live Databricks cluster. The optimized query must produce identical results and show a statistically meaningful speedup before it is accepted.

Examples

# Minimal — paste an inline query, discover everything automatically
/databricks-sql-autotuner "SELECT o.order_id, SUM(l.amount) FROM orders o JOIN lines l ON o.id = l.order_id GROUP BY o.order_id"

# From a file, cluster by ID
/databricks-sql-autotuner --cluster-id 0408-195905-abc --profile premium @queries/slow_report.sql

# From a file, cluster by name
/databricks-sql-autotuner --cluster-name my-cluster --profile premium @queries/slow_report.sql

# Optimize for speed, 5 benchmark runs