Welcome to the world of SQL Copilots
summary
Azure adds a Copilot for Azure SQL that demos about 40 skills (team cites ~55 total) and uses Azure OpenAI without fine‑tuning on customer data. It queries DMVs, Query Store and diagnostics, reads schema/keys (not row data), identifies a high‑CPU query executed ~16,000 times, suggests an index definition, spots blocking sessions, and generates runnable T‑SQL from natural language (pivot, rank). Preview is gated, permission‑aware, mostly read/advise today with execution gated by user consent.
AI-written summary · about
chapters (13)
- 0:00SQL Copilots intro
- 1:50What are SQL Copilots
- 3:56Azure Copilot & natural SQL
- 8:50Demo: DB size & performance
- 10:13High CPU & query tuning
- 12:20Query store troubleshooting
- 13:23Blocking processes with Copilot
- 15:51Natural SQL in editor
- 25:21How SQL Copilot works
- 32:40Copilot reliability & limits
- 33:21Pricing & availability
- 36:11Sign up for preview
- 42:14Q&A
related talks
45mWelcome to the world of SQL Copilots
09mHarness Microsoft Copilot in Azure for Azure SQL Database
40mAI Studio - Creating and scaling your custom copilots
47mUnlock potential on Azure with Microsoft Copilot
36mWhat's new with Copilot Studio
47mGitHub Copilot for SQL Development: Integrating the Power of AI into Database Development
45mThe era of the AI Copilot
45mBuilding a copilot: Azure AI Studio or Copilot Studio
