Data
Data warehouse
A data warehouse is a database built for analysis that collects data from many business systems in one place, so reports can combine them without slowing down the systems themselves.
How it works in practice
BigQuery, Snowflake, Redshift and Postgres used as an analytics database are common choices. Data arrives through ETL or ELT, is modelled into tables that match how the business thinks, and feeds dashboards, analysis and sometimes AI systems.
A small company does not always need one. If the questions are about one system, its own reporting may be enough. A warehouse earns its keep when people spend hours each week exporting from several systems into spreadsheets to answer basic questions, or when the numbers in different tools disagree.
A worked example
A 120-person distributor had sales in its ERP, margin in spreadsheets and stock in a warehouse system. A Postgres warehouse with nightly loads now answers margin by customer and product without anyone exporting anything.
Examples in this glossary describe typical builds for companies of 1 to 1,000 people. They are not named client results.















