Developing AI Applications with SQL AI: Ground to Cloud to Fabric
summary
SQL Server 2025 adds a native vector data type, T‑SQL functions to call models and generate embeddings, and DiskANN‑based vector indexes so semantic nearest‑neighbor queries run inside the database. A demo builds a multi‑agent app (Semantic Kernel + Data API Builder + Copilot Studio + REST model endpoints) that finds and summarizes a John Doe across CRM, policies and communications in about 3–5 minutes. Declarative create external model and an MCP server enable on‑prem, cloud or Fabric model hosting.
AI-written summary · about
chapters (8)
- 0:00AI and SQL integration
- 10:26Building AI applications
- 16:49Vector search in SQL
- 22:29AI agents and systems
- 37:43Model Context Protocol in SQL
- 41:17Enterprise apps with AI
- 46:15SQL Server 2025 AI features
- 55:17Q&A
related talks
46mPower AI apps and develop rich experiences with Azure SQL Database
50mBuilding AI ready applications
01h 16mAzure SQL ❤️ Open AI
36mFuture Data Architecture Enabled by New AI Technology
01h 12mBuilding Scalable GenAI Apps Using Semantic Kernel and .NET Aspire
01h 05mKeynote: Building Applications with the Modern SQL Using AI and Microsoft Fabric
50mHow to design and build AI applications with vector search using Azure OpenAI & Azure Cosmos DB
45mForget SQL! Create your own AI Data Analyst