# Converting database objects with generative AI
<a name="schema-conversion-convert.databaseobjects"></a>

The DMS Schema Conversion with generative AI feature streamlines the database migration process by offering recommendations to help you convert previously unconverted code objects that typically require complex manual conversion. This feature is available for Oracle to PostgreSQL/Aurora PostgreSQL, SQL Server to PostgreSQL/Aurora PostgreSQL and SAP ASE (Sybase ASE) to PostgreSQL/Aurora PostgreSQL conversions. You can convert an entire database schema or individual database schema objects.

**Warning**  
Generative AI features in DMS Schema Conversion use cross-region inference. For more information, see [Cross-region inference in DMS Schema Conversion](CHAP_Security.DataProtection.CrossRegionInference.md#CHAP_Security.DataProtection.CrossRegionInference.SchemaConversion).

To convert your source database objects with generative AI, follow steps 1 to 6 in [Converting database objects ](schema-conversion-convert.md#schema-conversion-convert-steps) then continue with one of these two methods:
+ Method 1: From the **Actions** menu, select **Convert**. In the conversion dialog box that appears, enable the **Convert schema with Generative AI** option and click **Convert**.
+ Method 2: Click ****Convert schema with Generative AI**** in the top right corner. In the conversion dialog box, ensure the option is enabled and click **Convert**.

To manually adjust this setting at any time in DMS Schema Conversion console:
+ Navigate to the **Settings** tab.
+ In the **Conversion settings section**, enable the **Generative AI** option to approve the use of generative AI.

**Note**  
Supported regions:  
Asia Pacific (Tokyo) (ap-northeast-1)
Asia Pacific (Osaka) (ap-northeast-3)
Asia Pacific (Sydney) (ap-southeast-2)
Canada (Central) (ca-central-1)
Europe (Frankfurt) (eu-central-1)
Europe (Stockholm) (eu-north-1)
Europe (Ireland) (eu-west-1)
Europe (London) (eu-west-2)
Europe (Paris) (eu-west-3)
US East (N. Virginia) (us-east-1)
US East (Ohio) (us-east-2)
US West (Oregon) (us-west-2)

**Note**  
Supported conversion paths:  
Oracle to Amazon RDS for PostgreSQL
Oracle to Amazon Aurora PostgreSQL
Microsoft SQL Server to Amazon RDS for PostgreSQL
Microsoft SQL Server to Amazon Aurora PostgreSQL
SAP ASE (Sybase ASE) to Amazon RDS for PostgreSQL
SAP ASE (Sybase ASE) to Amazon Aurora PostgreSQL
IBM Db2 for Linux, UNIX and Windows (LUW) to Amazon RDS for PostgreSQL
IBM Db2 for Linux, UNIX and Windows (LUW) to Amazon Aurora PostgreSQL
IBM Db2 for z/OS to Amazon RDS for PostgreSQL
IBM Db2 for z/OS to Amazon Aurora PostgreSQL

## Scope of Generative AI conversion
<a name="schema-conversion-convert.databaseobjects.genai"></a>

Generative AI-assisted schema conversion focuses on specific SQL elements with designated action items. All other SQL elements are converted using default rule-based approaches. The SQL elements within the extended scope of Generative AI conversion include:


****  
[\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/dms/latest/userguide/schema-conversion-convert.databaseobjects.html)

### Limitations
<a name="schema-conversion-convert.databaseobjects.limitations"></a>

The Converting database objects with generative AI feature has the following limitations:
+ Database endpoints supporting generative AI conversion are not visible in the AWS Console. You can view them only by exporting the assessment report as a PDF or CSV file.
+ As a probabilistic system, generative AI-assisted Schema Conversion may not achieve 100 percent accuracy in all conversions. It can also produce different results for the same SQL statements over a period of time. You must review and validate all conversion outputs.
+ Generative AI conversion is not supported for:
  + DEFAULT constraint in a table
  + DEFAULT value for a function or procedure parameter
  + COMPUTE COLUMN in a table
  + TRIGGER
  + COLUMN DATA TYPE
  + Dynamic SQL
  + INDEX
  + CONSTRAINT
+ If the source statement is converted with multiple action items and at least one action item is processed using generative AI, then all Action Items are replaced by one action item 5444 on a target for Oracle and 7744 for Microsoft SQL Server. The action item 9997 is an exception that is saved after processed using generative AI.

**Warning**  
Conversion using generative AI takes longer than basic conversion.

Every AWS account have a per-minute quota limiting the number of SQL statements that can be converted using generative AI. Statements exceeding this limit are queued for processing in subsequent minutes. The quota is as follows:


****  

| Region | SQL Statements per AWS account per minute | 
| --- | --- | 
|  Asia Pacific (Tokyo) (ap-northeast-1) Asia Pacific (Osaka) (ap-northeast-3) Asia Pacific (Sydney) (ap-southeast-2) Canada (Central) (ca-central-1) Europe (Stockholm) (eu-north-1) Europe (Ireland) (eu-west-1) Europe (London) (eu-west-2) Europe (Paris) (eu-west-3) US East (Ohio) (us-east-2)  | Up to 24 statements | 
|  Europe (Frankfurt) (eu-central-1) US East (N. Virginia) (us-east-1) US West (Oregon) (us-west-2)  | Up to 40 statements |