Accelerate AppDev with Database 23ai
summary
Oracle Database 23ai adds a native vector column type and a vector distance function so applications run similarity search directly in SQL, avoiding external vector stores and syncing. Embedding models (examples include image models that produce 1,024-dimensional vectors) can run against in‑database data, and generative features produce AI-generated SQL and support retrieval‑augmented generation. Claim: developers can add vector search plus business WHERE filters in minutes with no ML expertise and no separate vector DB.
AI-written summary · about
chapters (12)
- 0:00Introduction to AI
- 0:46Understanding vector data
- 2:22Vectors vs traditional arrays
- 4:04Visualizing similarity search
- 5:26Vector similarity in practice
- 6:55Vector databases explained
- 8:26Combining business and AI data
- 10:11Vector search in Oracle
- 12:05SQL for AI search
- 13:27Generative AI fundamentals
- 15:22Retrieval augmented generation
- 16:41Summary and conclusion
related talks
14mBring AI to Your Data
24mFuture of Data and AI
12mTackle Any Workload with Oracle Database, AI, and Multicloud
33mBringing Vector Search to OLTP
18mData Superpowers with Oracle Database 23ai
36mUnleashing the Power of Vector Databases for AI-Driven Applications
01h 21mKeynote on the "AI for Data" Revolution
48mOracle Database Directions: Innovate, Scale, and Sleep Easier