Data Lake

A data lake stores raw data — structured, semi-structured, and unstructured (logs, JSON, images) — in cheap object storage at massive scale, with schema applied when you read it (schema-on-read) rather than when you write it. It trades the structure of a warehouse for flexibility.

Worked example: dump raw event logs, clickstreams, and exports into the lake now, and decide how to structure and query them later — useful when you do not yet know every question you will ask. Gotcha: without governance a data lake becomes a “data swamp” — undocumented, inconsistent, unqueryable — which is why the “lakehouse” pattern adds table formats and schemas on top of lake storage to get warehouse-like reliability.