SQL Server 2025: AI-Ready Wherever You Run It
Bob Ward explains what “AI-ready” means in SQL Server 2025, focusing on native vector capabilities, embedding generation, and RAG patterns you can implement directly with T-SQL across local and cloud AI model deployments.
Overview
The video walks through SQL Server 2025 capabilities aimed at building AI-enabled applications wherever SQL Server runs (on-premises, in the cloud, and in sovereign environments).
AI-ready database capabilities
- Native vector capabilities for AI scenarios.
- Embedding generation support.
- Retrieval-augmented generation (RAG) application patterns implemented using T-SQL.
- Options to use local AI models or cloud-hosted AI models, depending on deployment constraints.
Developer and application features
- Optimized locking behavior (demonstrated in the session).
- Native JSON support.
- Regular expression support.
- Change event streaming for application integration.
- Data API builder for exposing data to applications.
Platform, operations, and integration topics
- Security considerations tied to application examples.
- Disaster recovery considerations.
- Microsoft Fabric mirroring integration.
- SQL Server on Azure Local, including an architecture designed to keep data, applications, and AI within a sovereign environment.