Keynote Review
summary
Databricks claims a Lakehouse 3.0 that ends format wars via Delta Lake, Delta Sharing and a Unity Catalog control plane, pushes a federated marketplace and touts Mosaic ML acquisition plus a GPU stockpile. Key numbers: two exabytes processed per day and 1,500 transformers running in-platform (up from 400). Positioning targets multicloud with deep Microsoft ties and JP Morgan as a testimonial; core pitch is keeping IP out of public LLMs.
AI-written summary · about
chapters (9)
- 0:00Opening the cube coverage
- 1:06Open source and unified data
- 3:05Multi-cloud strategy
- 5:26LLM tools and marketplace
- 7:23The data developer era
- 9:28AI hype vs reality
- 11:16AI ethics and IP concerns
- 13:27Adoption and cloud training
- 16:17Keynote analysis and wrap
