Merge pull request #53987 from JulianePadrao/update-build-ai-sql

Revise GitHub Copilot section in SSMS documentation for clarity and d…

Author: v-dirichards

learn-pr/wwl-data-ai/build-ai-solutions-sql-server/2-copilot.yml

@@ -1,9 +1,9 @@
 ### YamlMime:ModuleUnit
 uid: learn-wwl.build-ai-solutions-sql-server.copilot
-title: Use Copilot in SQL Server Management Studio
+title: Use GitHub Copilot in SQL Server Management Studio
 metadata:
-  title: Use Copilot in SQL Server Management Studio
-  description: "Learn how to use Copilot in SQL Server Management Studio"
+  title: Use GitHub Copilot in SQL Server Management Studio
+  description: "Learn how to use GitHub Copilot in SQL Server Management Studio"
   ms.date: 10/20/2025
   author: JulianePadrao
   ms.author: jupadrao

learn-pr/wwl-data-ai/build-ai-solutions-sql-server/includes/2-copilot.md

@@ -1,37 +1,44 @@
-[SQL Server Management Studio (SSMS)](/sql/ssms/sql-server-management-studio-ssms?azure-portal=true) supports two AI-powered coding assistants that help you write, optimize, and troubleshoot [T-SQL](/sql/t-sql/language-reference?azure-portal=true) code. [GitHub Copilot](/ssms/github-copilot/overview?azure-portal=true) provides general coding assistance across multiple languages, while [Copilot](/ssms/copilot/copilot-in-ssms-overview?azure-portal=true) offers database-aware suggestions specific to your connected SQL Server instance.
+[SQL Server Management Studio (SSMS)](/sql/ssms/sql-server-management-studio-ssms?azure-portal=true) includes [GitHub Copilot](/ssms/github-copilot/overview?azure-portal=true), an AI-powered assistant that helps you write, optimize, and troubleshoot [T-SQL](/sql/t-sql/language-reference?azure-portal=true) code. GitHub Copilot provides code completions in the query editor when writing T-SQL, and also provides a chat where you can use natural language to get help writing queries, ask questions about your database or environment, and get help with SQL issues. You should always use the latest release of SSMS (currently SSMS 22) as that is the GA release.
 
-Both tools use large language models to accelerate database development. GitHub Copilot excels at code completion and pattern recognition across your entire codebase, while Copilot leverages your database schema, indexes, and statistics to provide contextually aware T-SQL assistance.
+GitHub Copilot allows you to select a model, and available models vary based on your subscription. GitHub Copilot offers code completion and leverages your connection and database schema to provide context to the model and improve responses.
 
-These AI assistants make database development more accessible and improve productivity for both new and experienced developers. Whether you're writing queries, debugging code, or optimizing performance, AI-powered assistance helps you focus on solving business problems rather than remembering syntax.
+GitHub Copilot makes database development more accessible and improves productivity for both new and experienced developers. Whether you're writing queries, troubleshooting issues with your code, or optimizing performance, GitHub Copilot helps you focus on solving business problems rather than remembering syntax.
 
-## Use GitHub Copilot in SSMS
+## Understand GitHub Copilot in SQL Server Management Studio
 
-[GitHub Copilot in SSMS](/ssms/github-copilot/overview?azure-portal=true) brings AI-powered code completion and generation to your database development workflow. As a general-purpose coding assistant, GitHub Copilot helps you write T-SQL faster by suggesting code as you type and generating entire queries from natural language comments.
+[GitHub Copilot in SSMS](/ssms/github-copilot/overview?azure-portal=true) brings AI-powered code completion, generation, and chat to your database development workflow. GitHub Copilot helps you write T-SQL faster by suggesting code as you type, generating entire queries from natural language comments, and providing a chat experience where you can ask questions about your database and get help with SQL issues.
 
-### Understand GitHub Copilot capabilities
+### Explore GitHub Copilot capabilities
 
 GitHub Copilot provides several features for SQL Server development:
 
 - **Inline code completion:** Real-time suggestions as you write T-SQL code
 - **Natural language to code:** Write comments describing your intent, get working queries
 - **Multi-line suggestions:** Generate entire query blocks, procedures, or functions
-- **Pattern recognition:** Learn from your coding style and common patterns in your workspace
-- **Context-aware:** Considers the code you've already written in the current file
-- **Multi-language support:** Works with T-SQL, PowerShell, Python, and other languages used in database projects
+- **Context aware:** Understands the schema of the database to which you're connected
 
-GitHub Copilot analyzes your current file, open tabs, and workspace to provide relevant suggestions. While it doesn't connect directly to your database, it recognizes T-SQL syntax patterns and common database development practices.
+GitHub Copilot also offers these features via its chat:
+
+- **Schema-aware suggestions:** Knows your exact table structures, columns, and data types
+- **Query optimization:** Analyzes queries and suggests performance improvements
+- **Natural language queries:** Converts questions into T-SQL based on the database to which you're connected
+- **Explain queries:** Describes what existing queries do in plain language
+- **Fix errors:** Provides context-aware suggestions when queries fail
+- **Index recommendations:** Suggests indexes based on the missing index suggestions in the DMVs or in a query plan
+
+GitHub Copilot also provides assistance troubleshooting query and database performance, along with answering questions about database maintenance, best practices, database and server configuration, and more.
 
 ### Enable GitHub Copilot in SSMS
 
 To use GitHub Copilot in SQL Server Management Studio:
 
-1. Ensure you have [SSMS 19.3](/ssms/download-sql-server-management-studio-ssms?azure-portal=true) or later installed
-2. Install the [GitHub Copilot extension](https://docs.github.com/en/copilot/how-tos/set-up/install-copilot-extension) from the Extensions menu
+1. Ensure you have [SSMS 22](/ssms/download-sql-server-management-studio-ssms?azure-portal=true) installed
+2. Install GitHub Copilot by selecting the **AI Assistance** workload in the [Visual Studio Installer](/ssms/github-copilot/installation-state?azure-portal=true)
 3. Sign in with your GitHub account that has an active [GitHub Copilot subscription](https://github.com/features/copilot/plans?azure-portal=true)
-4. Configure your preferences in **Tools > Options > GitHub Copilot**
+4. Configure your preferences in **Tools > Options > GitHub > Copilot** and **Tools > Options > Text Editor > Inline Suggestions**
 
 > [!NOTE]
-> GitHub Copilot requires a paid subscription (Individual, Business, or Enterprise). Some organizations provide GitHub Copilot for their developers as part of their development tools license.
+> GitHub Copilot requires a subscription (Individual, Business, or Enterprise). Some organizations provide GitHub Copilot for their developers as part of their development tools license. A GitHub Copilot subscription can also be used with Visual Studio and VS Code.
 
 ### Write code with inline suggestions
 
@@ -40,13 +47,13 @@ GitHub Copilot provides suggestions as you type:
 ```sql
 -- Start typing a query
 SELECT c.CustomerName, 
--- Copilot suggests: c.Email, c.Phone, c.City
+-- GitHub Copilot suggests: c.Email, c.Phone, c.City
 -- Press Tab to accept
 
 -- Or write a comment describing what you need:
 -- Get total sales by product category for last quarter
 
--- Copilot generates:
+-- GitHub Copilot generates:
 SELECT 
     p.Category,
     SUM(od.Quantity * od.UnitPrice) AS TotalSales
@@ -58,7 +65,7 @@ GROUP BY p.Category
 ORDER BY TotalSales DESC;
 ```
 
-The gray ghost text shows Copilot's suggestions. Press **Tab** to accept or **Esc** to dismiss.
+The gray ghost text shows GitHub Copilot's suggestions. Press **Tab** to accept or **Esc** to dismiss.
 
 ### Generate stored procedures and functions
 
@@ -96,7 +103,7 @@ GitHub Copilot helps write queries to explore your database structure:
 ```sql
 -- Show all foreign key relationships for Orders table
 
--- Copilot suggests:
+-- GitHub Copilot suggests:
 SELECT 
     fk.name AS ForeignKeyName,
     OBJECT_NAME(fk.parent_object_id) AS TableName,
@@ -109,9 +116,9 @@ WHERE OBJECT_NAME(fk.parent_object_id) = 'Orders'
    OR OBJECT_NAME(fk.referenced_object_id) = 'Orders';
 ```
 
-### Learn from your coding patterns
+### Apply coding patterns of the current editor
 
-GitHub Copilot adapts to your style over time. If you consistently use specific naming conventions, formatting, or patterns, Copilot incorporates them into suggestions:
+GitHub Copilot follows the coding patterns in your current editor. If you use specific naming conventions, formatting, or patterns, GitHub Copilot incorporates them into suggestions:
 
 ```sql
 -- If you typically write your queries like this:
@@ -122,7 +129,7 @@ SELECT
 FROM Customers AS c
 WHERE c.IsActive = 1;
 
--- Copilot learns your style and suggests:
+-- GitHub Copilot follows your style and suggests:
 SELECT 
     o.OrderID
     ,o.OrderDate
@@ -131,7 +138,7 @@ FROM Orders AS o
 WHERE o.Status = 'Completed';
 ```
 
-This personalization makes suggestions feel more natural and reduces the need for manual formatting.
+This consistency makes suggestions feel more natural and reduces the need for manual formatting.
 
 ### Apply GitHub Copilot best practices
 
@@ -140,40 +147,20 @@ Maximize GitHub Copilot's effectiveness:
 | Practice | Description |
 |----------|-------------|
 | **Write descriptive comments** | Clear, specific comments produce better code suggestions |
-| **Use meaningful names** | Descriptive table and column names help Copilot understand context |
+| **Use meaningful names** | Descriptive table and column names help GitHub Copilot understand context |
 | **Break down complex tasks** | Write multiple comments for complex procedures rather than one large block |
 | **Review all suggestions** | Always validate generated code for correctness and security |
 | **Provide examples** | Include sample data or expected output in comments |
-| **Keep context nearby** | Open related files or schemas to give Copilot more context |
-| **Iterate on suggestions** | If the first suggestion isn't quite right, try rephrasing your comment |
-
-## Use Copilot in SSMS
-
-[Copilot in SSMS](/ssms/copilot/copilot-in-ssms-overview?azure-portal=true) provides database-aware AI assistance that connects directly to your SQL Server instance. Unlike GitHub Copilot, Copilot has access to your database schema, indexes, statistics, and metadata, enabling more contextually accurate T-SQL suggestions.
-
-:::image type="content" source="../media/ssms.png" alt-text="Screenshot showing GitHub Copilot in SQL Server Management Studio with AI-powered code completion." lightbox="../media/ssms.png" border="false":::
-
-### Understand Copilot capabilities
-
-Copilot offers database-specific features:
-
-- **Schema-aware suggestions:** Knows your exact table structures, columns, and data types
-- **Query optimization:** Analyzes execution plans and suggests performance improvements
-- **Natural language queries:** Converts questions into T-SQL based on your actual database
-- **Explain queries:** Describes what existing queries do in plain language
-- **Fix errors:** Provides context-aware suggestions when queries fail
-- **Index recommendations:** Suggests indexes based on your database statistics
-
-Because Copilot connects to your database, it generates code that works with your specific schema without modification.
+| **Iterate on suggestions** | If the first suggestion isn't right, try rephrasing your comment |
 
 ### Query with natural language
 
 Ask questions about your data using plain English:
 
 ```sql
--- Ask Copilot: "Show customers who placed more than 5 orders last month"
+-- Ask GitHub Copilot: "Show customers who placed more than 5 orders last month"
 
--- Copilot generates (using your actual schema):
+-- GitHub Copilot generates (using your actual schema):
 SELECT 
     c.CustomerID,
     c.CustomerName,
@@ -188,11 +175,11 @@ HAVING COUNT(o.OrderID) > 5
 ORDER BY OrderCount DESC;
 ```
 
-Copilot knows your exact table and column names, ensuring the query works immediately.
+Because GitHub Copilot connects to your database, it generates code that works with your specific schema.
 
 ### Optimize query performance
 
-Copilot analyzes queries and suggests improvements:
+GitHub Copilot analyzes queries and suggests improvements:
 
 ```sql
 -- Select a slow query and ask: "How can I make this faster?"
@@ -205,7 +192,7 @@ LEFT JOIN OrderDetails od ON o.OrderID = od.OrderID
 WHERE o.OrderDate >= '2024-01-01'
 GROUP BY c.CustomerName;
 
--- Copilot suggests:
+-- GitHub Copilot suggests:
 -- 1. Change LEFT JOIN to INNER JOIN (WHERE filters out NULLs anyway)
 -- 2. Add covering index: CREATE INDEX IX_Orders_Date ON Orders(OrderDate, CustomerID, OrderID)
 -- 3. Consider indexed view for frequent aggregations
@@ -222,7 +209,7 @@ OPTION (RECOMPILE);
 
 ### Understand and fix errors
 
-When queries fail, Copilot explains the issue and provides fixes:
+When queries fail, GitHub Copilot explains the issue and provides fixes:
 
 ```sql
 -- Query with error:
@@ -231,9 +218,9 @@ FROM Orders;
 
 -- Error: Column 'Orders.CustomerName' is invalid in the select list
 
--- Ask Copilot: "What's wrong with this query?"
+-- Ask GitHub Copilot: "What's wrong with this query?"
 
--- Copilot explains and fixes:
+-- GitHub Copilot explains and fixes:
 /*
 The error occurs because CustomerName is not in an aggregate function or GROUP BY clause.
 When using aggregate functions like SUM(), all non-aggregated columns must be grouped.
@@ -245,19 +232,8 @@ FROM Orders
 GROUP BY CustomerName;
 ```
 
-### Choose the right tool
-
-Both tools offer value for SQL Server development:
-
-| Feature | GitHub Copilot | Copilot |
-|---------|----------------|-------------------|
-| **Database schema awareness** | No | Yes |
-| **Multi-language support** | Yes | T-SQL focus |
-| **Query optimization** | Pattern-based | Database-specific |
-| **Works outside SSMS** | Yes | No |
-| **Requires database connection** | No | Yes |
-| **Subscription** | GitHub Copilot | Copilot |
+### Get the most from GitHub Copilot
 
-Use **GitHub Copilot** for general T-SQL development, multi-language projects, and code completion across your entire codebase. Use **Copilot** for database-specific tasks where schema awareness, query optimization, and direct database interaction provide better results.
+The GitHub Copilot chat window works without a database connection, but it's much more valuable when you have a query editor open and connected to a database. Use GitHub Copilot for general T-SQL development, code completion in the editor, understanding your database and server, query optimization, and more.
 
-Many developers use both: GitHub Copilot for day-to-day coding and Copilot when working with specific database instances. Together, these AI assistants accelerate database development while maintaining code quality and security standards.
+Developers can use GitHub Copilot for day-to-day coding and to answer specific database questions, helping accelerate database development while maintaining code quality and leveraging best practices.