Azure SQL Dev Corner
Voices from the Azure SQL PM Team, focusing on development and developers
Latest posts
Considering NL2SQL? Should your database really be the prompt? How can SQL MCP Server help?
You’ve probably experienced both of these, perhaps at the same time. First, that desire to let an agent get at your data. It’s driven by simplification and better experiences for the user and for you: fewer screens, fewer queries, fewer reports, and less code overall. Second, and perhaps more importantly, that unrelenting reluctance and reticence against all of it. The voice in your head that makes you seriously uneasy. Your reputation is at stake, yes, but to your enterprise, this might be existential. It has a name: NL2SQL. NL2SQL, or Natural Language to SQL, is where an AI agent turns a natural language ...
Microsoft SQL Security Across the MAESTRO Stack: Building Secure Agentic AI with Defense-in-Depth
Artificial Intelligence is evolving rapidly. What began as simple prompt-and-response systems is now transforming into fully autonomous, agentic AI architectures capable of reasoning, orchestrating tools, interacting with enterprise data, and invoking external systems dynamically. While these capabilities unlock enormous business potential, they also introduce an entirely new category of security challenges. Organizations are no longer asking only: “How do we build AI systems?” They are now asking: “How do we build AI systems securely, responsibly, and with governance built into every layer?” This ...
Generate Embeddings Function and External Model Object Support Are Now Generally Available in Azure SQL
We are excited to announce the General Availability (GA) of AI_GENERATE_EMBEDDINGS and CREATE EXTERNAL MODEL in Azure SQL Database and Azure SQL Managed Instance. These two T-SQL features — CREATE EXTERNAL MODEL and AI_GENERATE_EMBEDDINGS — work together as a single, integrated pipeline for generating vector embeddings directly from T-SQL. No data movement, no external orchestration, no application-layer pipeline required. EXTERNAL MODEL object defines where to get embeddings from and how to authenticate — registered once, reused everywhere. AI_GENERATE_EMBEDDINGS calls the external model to generate the ve...
SQL MCP Server as an App Service
Run SQL MCP Server on Azure App Service without containers. This walkthrough uses Data API builder to configure authentication, expose MCP, REST, and GraphQL endpoints, and deploy as code.
Azure Data Studio is retired: Move your Azure SQL workflow to VS Code in 10 minutes
Azure Data Studio (ADS) retired on February 6, 2025, and support ended on February 28, 2026. The recommended path forward is Visual Studio Code with the MSSQL extension. If you used ADS daily, this guide gets you productive quickly in VS Code. The focus is on importing your existing setup, restoring familiar shortcuts like F5, and getting SQL Database Projects working so you can build and publish schema changes confidently from the editor. TL; DR Why this matters for Azure SQL developers Running queries is only part of the job. Most teams need repeatable workflows for schema change review, ...
The Polyglot tax – Part 4
The Agent-Ready Database: Security, Backup, and MCP Part 4 of 4 – The Multi-Model Database Series This is the final post in a four-part series on multi-model databases in SQL Server 2025 and Azure SQL - exploring how the optimizer, storage engine, and security layer treat each data model as a first-class citizen under one roof. In Part 1: The Polyglot Tax, we described the trajectory: you spin up a database, point an agent at it, and start building fast. The complexity comes later - JSON, graph, vectors, analytics - and each new requirement tempts you to spin up another database. In Part 2: When JSON Met Gra...
Introducing SQL MCP Server
SQL MCP Server gives enterprises a secure, feature-rich way to enable agents to access data. This is accomplished without exposing the schema, risking consistency, or relying on fragile natural language parsing. SQL MCP Server is a feature of Data API builder, so deployments have a proven entity abstraction system, RBAC security at the API layer with Azure Key Vault integration, custom OAuth and Microsoft Entra support, first-level and second-level caching with integration with Redis and Azure Managed Redis, and complete instrumentation and telemetry with integration with Azure Log Analytics, Application Insights...
The Polyglot tax – Part 3
Vectors, Analytics, and the End of ETL Part 3 of 4 - The Multi-model Database Series In Part 1: The Polyglot Tax we laid out the fraud detection scenario: a transaction comes in, and before you approve or deny it you need five checks. Order history (relational). Device fingerprint (JSON). Connections to fraud rings (graph). Transactions that look like this one (vector similarity). Statistical baselines across millions of rows (analytics). Five data access patterns, five databases in the polyglot model, five sets of everything that can break. In Part 2: When JSON Met Graph we handled the first three. We store...
SQL code analysis in VS Code: Configure rules without editing your project file
SQL code analysis has been part of the SSDT workflow for a long time. Before deploying a schema change, you could run a set of static analysis rules against your project to catch potential issues, things like missing primary keys, deprecated syntax, or objects that could break under certain compatibility levels. It was one of those SSDT features that teams quietly relied on without thinking much about it. When developers started moving their SQL Database Projects workflow to VS Code, code analysis came with them. But configuring which rules to enable or disable required editing the .sqlproj file directly, not a ...