Data Warehouse Query Cost Optimization Analysis Builder

You are an analytics engineering lead. Build a data warehouse query cost optimization analysis for:

**Warehouse:** [BigQuery, Snowflake, Redshift, Databricks, Postgres analytics replica]
**Billing model:** [bytes scanned, warehouse credits, compute clusters, slots, storage plus compute]
**Time period:** [date range]
**Available logs:** [query text, user, role, warehouse, dashboard, model, bytes scanned, duration, credits/cost, schedule time, success/failure]
**Workloads:** [BI dashboards, dbt models, notebooks, reverse ETL, ad hoc analysis, product analytics]
**Known pain:** [monthly bill spike, slow dashboards, runaway jobs, duplicate models, leadership cost target]
**Constraints:** [must preserve SLAs, compliance retention, executive dashboards, data freshness]

Produce a cost analysis with:

1. **Spend baseline** — daily/weekly cost trend, forecasted monthly run rate, and change vs prior period.
2. **Top cost drivers** — queries, dashboards, models, users, roles, warehouses, schedules, and tables.
3. **Cost per business object** — estimate cost per dashboard, data product, team, or KPI where possible.
4. **SQL and modeling anti-patterns** — unbounded dates, SELECT *, cross joins, repeated CTE scans, excessive refreshes, missing clustering/partition pruning, oversized warehouses.
5. **Freshness vs cost tradeoff** — which jobs can run less often, incrementally, or only on business days.
6. **Quick wins** — low-risk changes with estimated savings and owners.
7. **Structural changes** — materialized views, incremental models, aggregate tables, semantic layer caching, warehouse sizing, retention policies.
8. **SQL/log queries** — example queries to reproduce the analysis.
9. **Risk assessment** — what could break and how to roll back safely.
10. **Executive memo** — savings target, timeline, dependencies, and next review cadence.

Do not recommend blind cost cutting; preserve trusted reporting and document expected freshness changes.