Finds where Snowflake credits burn: attributes 1M queries to cost patterns in 3.7s via literal-stripping query fingerprints, detects spills, full scans, and idle warehouses, and estimates right-sizing savings with the math shown. Runs fully offline against a committed DuckDB replica of ACCOUNT_USAGE; same SQL runs on a real account.
python cli sql snowflake data-engineering query-optimization observability cost-optimization finops duckdb healthcare-analytics account-usage warehouse-sizing
-
Updated
Aug 11, 2026 - Python