Panel Discussion
summary
Panel explores practical pain and fixes: thousands of messy models (about 8,000 dbt models at one company, a cited case of 39,000 dbt models) driving cost and sprawl on Snowflake, teams embedding engineers in product squads to refactor and reuse dimensional/fact tables, and adopting data contracts to stop schema breakages. Semantic layers help consolidate definitions but only bandage upstream quality problems; expect more unstructured text work, metadata-driven automation, and AI-assisted documentation and cataloging.
AI-written summary · about
chapters (14)
- 0:00Welcome and panelists
- 1:06Data modeling defined
- 3:32Purpose and challenges
- 7:23Usability and performance
- 9:10Scale and sprawl management
- 12:42Semantic layers and upstream issues
- 16:11Future beyond Kimball
- 18:40Data contracts and feedback
- 21:48Q&A on complexity
- 26:19Improve data culture
- 30:10High-performing data teams
- 32:17AI in data platforms
- 36:31Advice for new engineers
- 38:59Closing
related talks
49mPanel discussion - Building Data Products
31mFuture of data engineering: what's in store?
49mData Modelling Showdown: Unravelling the Mysteries of Different Data Modelling Approaches
33mPanel Discussion
34mDimensional Data Modeling in the Modern Era
27mDataEngBytes 2023 - BRI-T1-07 - Joe Reis
22mData Modelling is not a one man job!
31mPanel Discussion