Data warehousing is the practice of consolidating data from multiple operational and external sources into a central, integrated, subject-oriented repository optimised for query and analysis rather than transaction processing. A data warehouse stores historical, cleansed and conformed data structured for reporting and decision support, typically populated through extract-transform-load pipelines and queried using online analytical processing. It provides the persistent analytical substrate on which business intelligence, dashboards and downstream analytics are built.
Overview
- A data warehouse is a subject-oriented, integrated, time-variant and non-volatile collection of data built to support management decision-making.
- Source data is ingested through ETL or ELT pipelines that cleanse, conform and load it into a query-optimised schema.
- Warehouses commonly use dimensional models (star and snowflake schemas) to make analytical queries fast and intuitive for business users.
Key aspects
- Integration: heterogeneous operational sources are unified into a single conformed model with consistent definitions.
- Historical retention: data is stored over time to support trend analysis and point-in-time reporting.
- Read-optimised: storage and indexing favour large analytical scans rather than transactional inserts and updates.
- Governance: master data management and data quality controls underpin trustworthy analytics.
Applications
- Enterprise reporting and executive dashboards.
- Self-service business intelligence and ad-hoc analysis.
- Feeding predictive analytics and machine-learning feature stores.
- Regulatory and financial reporting requiring auditable history.