data-warehouse

Use this skill for 大数据/数仓（DW）建设与运维：从数据接入到指标交付与治理。

Defaults / assumptions to confirm

• Warehouse tech: BigQuery/Snowflake/Redshift/Hive/ClickHouse/etc.
• Orchestration: Airflow/Dagster/Argo/dbt
• Ingestion: CDC vs batch, streaming (Kafka) vs file
• Data consumers: BI dashboards, product analytics, ML features

Core outputs

• Warehouse architecture (sources → staging → warehouse → marts)
• Data model (facts/dimensions) + metric definitions
• Pipeline plan (DAGs, schedules, dependencies, SLAs)
• Data quality plan (checks, thresholds, alerts)
• Cost/performance plan (partitioning, clustering, materialization)
• Governance plan (access control, PII handling, retention)