Building Tomorrow's Warehouse Today: Modern Data Warehouse Modelling Insights
summary
Modern warehouses combine multiple cloud services, use ELT not ETL pipelines to decouple extraction and transformations, and mix normalized and denormalized stores with virtualization or Delta Lake for transactional and analytic needs. Domain teams publish data products in a data mesh while a centralized data catalog provides automated metadata and data lineage (table/column/row). Streaming plus batch remains common, Data Vault and graph models solve history and relationships, and the recommendation is pragmatic: don’t keep raw noise forever, focus on business outcomes.
AI-written summary · about
chapters (13)
- 0:00Modern data warehousing intro
- 1:31Key data warehouse terms
- 9:59Streaming vs batch processing
- 12:31Natural language processing and machine learning in data warehouses
- 15:51Data governance and catalogs
- 17:20Data warehouse architectures
- 23:04Centralized vs decentralized data
- 24:29Choosing data modeling structures
- 28:24Traditional challenges and cloud solutions
- 31:15What data to keep
- 33:24Purpose of data warehouse
- 36:37Future of data warehousing
- 41:16Closing remarks
related talks
41mModern Data Architectures – What They Are and How to Implement?
52mData Warehouse, Data Lake, Data Mesh, Oh My: The History of Data Storage
31mData Warehouses Meet Data Lakes
27mThe Evolution of Data Architecture
01h 01mModern Data Warehousing Strategy
05mThe Future of Data Warehousing
36mIntroducing Data Mesh A Paradigm Shift In Analytical Data Management Architecture
26mLogical data fabric for building a unified data warehouse and data lake