What it means
Data lakes retain diverse raw and curated data, while warehouses structure data for governed analysis. Modern platforms may combine these patterns through lakehouse architectures.
Why it matters
Centralized analytics can create powerful business value and equally powerful concentration risk when classification, lineage, access, and retention are weak.
Practical focus
- Catalog data, ownership, lineage, and sensitivity
- Separate ingestion, transformation, and consumer permissions
- Monitor high-volume and unusual queries or exports
Common mistakes
- Treating a data lake as an unrestricted dumping ground
- Copying sensitive data into analytics without purpose review
- Using one shared service identity for every pipeline
Certification relevance
This subject appears in or supports the following certification bodies of knowledge: