You can use the Amazon Bedrock catalog source to extract metadata from a source system.
Amazon Bedrock is a fully managed service to build and scale generative AI applications without the need to manage any underlying infrastructure.
Extracted metadata
You can extract specific metadata from source systems that you connect to with an Amazon Bedrock catalog source.
Objects extracted
You can extract metadata of the following objects from an Amazon Bedrock source system:
•Agent
- Agent Alias
- Agent Version
▪ Action Group
▪ Action Group Function
•Amazon Bedrock Tag
•AWS Account
•Flow
- Flow Alias
- Flow Version
•Knowledge Base
- Managed Knowledge Base
- Structured Data Store Knowledge Base
- Unstructured Vector Store Knowledge Base
Note:
When you edit agents or flows in the Amazon Bedrock source system, the system maintains a working draft version before you publish it. In Data Governance and Catalog, working draft versions appear as DRAFT. Associated test aliases appear under generated system names, such as Agent Test Alias for agent aliases and TSTALIASID for flow aliases.
Compatible connectors
Before you configure a catalog source, you must connect to the Amazon Bedrock source system.
Use the Amazon Bedrock connector to connect to the Amazon Bedrock source system.
For information about configuring a connection, see Connections.