AgentsforBedrock / Client / create_knowledge_base

create_knowledge_base#

AgentsforBedrock.Client.create_knowledge_base(**kwargs)#

Creates a knowledge base that contains data sources from which information can be queried and used by LLMs. To create a knowledge base, you must first set up your data sources and configure a supported vector store. For more information, see Set up your data for ingestion.

Note

If you prefer to let Amazon Bedrock create and manage a vector store for you in Amazon OpenSearch Service, use the console. For more information, see Create a knowledge base.

  • Provide the name and an optional description.

  • Provide the Amazon Resource Name (ARN) with permissions to create a knowledge base in the roleArn field.

  • Provide the embedding model to use in the embeddingModelArn field in the knowledgeBaseConfiguration object.

  • Provide the configuration for your vector store in the storageConfiguration object.

See also: AWS API Documentation

Request Syntax

response = client.create_knowledge_base(
    clientToken='string',
    description='string',
    knowledgeBaseConfiguration={
        'type': 'VECTOR',
        'vectorKnowledgeBaseConfiguration': {
            'embeddingModelArn': 'string'
        }
    },
    name='string',
    roleArn='string',
    storageConfiguration={
        'opensearchServerlessConfiguration': {
            'collectionArn': 'string',
            'fieldMapping': {
                'metadataField': 'string',
                'textField': 'string',
                'vectorField': 'string'
            },
            'vectorIndexName': 'string'
        },
        'pineconeConfiguration': {
            'connectionString': 'string',
            'credentialsSecretArn': 'string',
            'fieldMapping': {
                'metadataField': 'string',
                'textField': 'string'
            },
            'namespace': 'string'
        },
        'rdsConfiguration': {
            'credentialsSecretArn': 'string',
            'databaseName': 'string',
            'fieldMapping': {
                'metadataField': 'string',
                'primaryKeyField': 'string',
                'textField': 'string',
                'vectorField': 'string'
            },
            'resourceArn': 'string',
            'tableName': 'string'
        },
        'redisEnterpriseCloudConfiguration': {
            'credentialsSecretArn': 'string',
            'endpoint': 'string',
            'fieldMapping': {
                'metadataField': 'string',
                'textField': 'string',
                'vectorField': 'string'
            },
            'vectorIndexName': 'string'
        },
        'type': 'OPENSEARCH_SERVERLESS'|'PINECONE'|'REDIS_ENTERPRISE_CLOUD'|'RDS'
    },
    tags={
        'string': 'string'
    }
)
Parameters:
  • clientToken (string) –

    A unique, case-sensitive identifier to ensure that the API request completes no more than one time. If this token matches a previous request, Amazon Bedrock ignores the request, but does not return an error. For more information, see Ensuring idempotency.

    This field is autopopulated if not provided.

  • description (string) – A description of the knowledge base.

  • knowledgeBaseConfiguration (dict) –

    [REQUIRED]

    Contains details about the embeddings model used for the knowledge base.

    • type (string) – [REQUIRED]

      The type of data that the data source is converted into for the knowledge base.

    • vectorKnowledgeBaseConfiguration (dict) –

      Contains details about the embeddings model that’sused to convert the data source.

      • embeddingModelArn (string) – [REQUIRED]

        The Amazon Resource Name (ARN) of the model used to create vector embeddings for the knowledge base.

  • name (string) –

    [REQUIRED]

    A name for the knowledge base.

  • roleArn (string) –

    [REQUIRED]

    The Amazon Resource Name (ARN) of the IAM role with permissions to invoke API operations on the knowledge base.

  • storageConfiguration (dict) –

    [REQUIRED]

    Contains details about the configuration of the vector database used for the knowledge base.

    • opensearchServerlessConfiguration (dict) –

      Contains the storage configuration of the knowledge base in Amazon OpenSearch Service.

      • collectionArn (string) – [REQUIRED]

        The Amazon Resource Name (ARN) of the OpenSearch Service vector store.

      • fieldMapping (dict) – [REQUIRED]

        Contains the names of the fields to which to map information about the vector store.

        • metadataField (string) – [REQUIRED]

          The name of the field in which Amazon Bedrock stores metadata about the vector store.

        • textField (string) – [REQUIRED]

          The name of the field in which Amazon Bedrock stores the raw text from your data. The text is split according to the chunking strategy you choose.

        • vectorField (string) – [REQUIRED]

          The name of the field in which Amazon Bedrock stores the vector embeddings for your data sources.

      • vectorIndexName (string) – [REQUIRED]

        The name of the vector store.

    • pineconeConfiguration (dict) –

      Contains the storage configuration of the knowledge base in Pinecone.

      • connectionString (string) – [REQUIRED]

        The endpoint URL for your index management page.

      • credentialsSecretArn (string) – [REQUIRED]

        The Amazon Resource Name (ARN) of the secret that you created in Secrets Manager that is linked to your Pinecone API key.

      • fieldMapping (dict) – [REQUIRED]

        Contains the names of the fields to which to map information about the vector store.

        • metadataField (string) – [REQUIRED]

          The name of the field in which Amazon Bedrock stores metadata about the vector store.

        • textField (string) – [REQUIRED]

          The name of the field in which Amazon Bedrock stores the raw text from your data. The text is split according to the chunking strategy you choose.

      • namespace (string) –

        The namespace to be used to write new data to your database.

    • rdsConfiguration (dict) –

      Contains details about the storage configuration of the knowledge base in Amazon RDS. For more information, see Create a vector index in Amazon RDS.

      • credentialsSecretArn (string) – [REQUIRED]

        The Amazon Resource Name (ARN) of the secret that you created in Secrets Manager that is linked to your Amazon RDS database.

      • databaseName (string) – [REQUIRED]

        The name of your Amazon RDS database.

      • fieldMapping (dict) – [REQUIRED]

        Contains the names of the fields to which to map information about the vector store.

        • metadataField (string) – [REQUIRED]

          The name of the field in which Amazon Bedrock stores metadata about the vector store.

        • primaryKeyField (string) – [REQUIRED]

          The name of the field in which Amazon Bedrock stores the ID for each entry.

        • textField (string) – [REQUIRED]

          The name of the field in which Amazon Bedrock stores the raw text from your data. The text is split according to the chunking strategy you choose.

        • vectorField (string) – [REQUIRED]

          The name of the field in which Amazon Bedrock stores the vector embeddings for your data sources.

      • resourceArn (string) – [REQUIRED]

        The Amazon Resource Name (ARN) of the vector store.

      • tableName (string) – [REQUIRED]

        The name of the table in the database.

    • redisEnterpriseCloudConfiguration (dict) –

      Contains the storage configuration of the knowledge base in Redis Enterprise Cloud.

      • credentialsSecretArn (string) – [REQUIRED]

        The Amazon Resource Name (ARN) of the secret that you created in Secrets Manager that is linked to your Redis Enterprise Cloud database.

      • endpoint (string) – [REQUIRED]

        The endpoint URL of the Redis Enterprise Cloud database.

      • fieldMapping (dict) – [REQUIRED]

        Contains the names of the fields to which to map information about the vector store.

        • metadataField (string) – [REQUIRED]

          The name of the field in which Amazon Bedrock stores metadata about the vector store.

        • textField (string) – [REQUIRED]

          The name of the field in which Amazon Bedrock stores the raw text from your data. The text is split according to the chunking strategy you choose.

        • vectorField (string) – [REQUIRED]

          The name of the field in which Amazon Bedrock stores the vector embeddings for your data sources.

      • vectorIndexName (string) – [REQUIRED]

        The name of the vector index.

    • type (string) – [REQUIRED]

      The vector store service in which the knowledge base is stored.

  • tags (dict) –

    Specify the key-value pairs for the tags that you want to attach to your knowledge base in this object.

    • (string) –

      • (string) –

Return type:

dict

Returns:

Response Syntax

{
    'knowledgeBase': {
        'createdAt': datetime(2015, 1, 1),
        'description': 'string',
        'failureReasons': [
            'string',
        ],
        'knowledgeBaseArn': 'string',
        'knowledgeBaseConfiguration': {
            'type': 'VECTOR',
            'vectorKnowledgeBaseConfiguration': {
                'embeddingModelArn': 'string'
            }
        },
        'knowledgeBaseId': 'string',
        'name': 'string',
        'roleArn': 'string',
        'status': 'CREATING'|'ACTIVE'|'DELETING'|'UPDATING'|'FAILED'|'DELETE_UNSUCCESSFUL',
        'storageConfiguration': {
            'opensearchServerlessConfiguration': {
                'collectionArn': 'string',
                'fieldMapping': {
                    'metadataField': 'string',
                    'textField': 'string',
                    'vectorField': 'string'
                },
                'vectorIndexName': 'string'
            },
            'pineconeConfiguration': {
                'connectionString': 'string',
                'credentialsSecretArn': 'string',
                'fieldMapping': {
                    'metadataField': 'string',
                    'textField': 'string'
                },
                'namespace': 'string'
            },
            'rdsConfiguration': {
                'credentialsSecretArn': 'string',
                'databaseName': 'string',
                'fieldMapping': {
                    'metadataField': 'string',
                    'primaryKeyField': 'string',
                    'textField': 'string',
                    'vectorField': 'string'
                },
                'resourceArn': 'string',
                'tableName': 'string'
            },
            'redisEnterpriseCloudConfiguration': {
                'credentialsSecretArn': 'string',
                'endpoint': 'string',
                'fieldMapping': {
                    'metadataField': 'string',
                    'textField': 'string',
                    'vectorField': 'string'
                },
                'vectorIndexName': 'string'
            },
            'type': 'OPENSEARCH_SERVERLESS'|'PINECONE'|'REDIS_ENTERPRISE_CLOUD'|'RDS'
        },
        'updatedAt': datetime(2015, 1, 1)
    }
}

Response Structure

  • (dict) –

    • knowledgeBase (dict) –

      Contains details about the knowledge base.

      • createdAt (datetime) –

        The time at which the knowledge base was created.

      • description (string) –

        The description of the knowledge base.

      • failureReasons (list) –

        A list of reasons that the API operation on the knowledge base failed.

        • (string) –

      • knowledgeBaseArn (string) –

        The Amazon Resource Name (ARN) of the knowledge base.

      • knowledgeBaseConfiguration (dict) –

        Contains details about the embeddings configuration of the knowledge base.

        • type (string) –

          The type of data that the data source is converted into for the knowledge base.

        • vectorKnowledgeBaseConfiguration (dict) –

          Contains details about the embeddings model that’sused to convert the data source.

          • embeddingModelArn (string) –

            The Amazon Resource Name (ARN) of the model used to create vector embeddings for the knowledge base.

      • knowledgeBaseId (string) –

        The unique identifier of the knowledge base.

      • name (string) –

        The name of the knowledge base.

      • roleArn (string) –

        The Amazon Resource Name (ARN) of the IAM role with permissions to invoke API operations on the knowledge base.

      • status (string) –

        The status of the knowledge base. The following statuses are possible:

        • CREATING – The knowledge base is being created.

        • ACTIVE – The knowledge base is ready to be queried.

        • DELETING – The knowledge base is being deleted.

        • UPDATING – The knowledge base is being updated.

        • FAILED – The knowledge base API operation failed.

      • storageConfiguration (dict) –

        Contains details about the storage configuration of the knowledge base.

        • opensearchServerlessConfiguration (dict) –

          Contains the storage configuration of the knowledge base in Amazon OpenSearch Service.

          • collectionArn (string) –

            The Amazon Resource Name (ARN) of the OpenSearch Service vector store.

          • fieldMapping (dict) –

            Contains the names of the fields to which to map information about the vector store.

            • metadataField (string) –

              The name of the field in which Amazon Bedrock stores metadata about the vector store.

            • textField (string) –

              The name of the field in which Amazon Bedrock stores the raw text from your data. The text is split according to the chunking strategy you choose.

            • vectorField (string) –

              The name of the field in which Amazon Bedrock stores the vector embeddings for your data sources.

          • vectorIndexName (string) –

            The name of the vector store.

        • pineconeConfiguration (dict) –

          Contains the storage configuration of the knowledge base in Pinecone.

          • connectionString (string) –

            The endpoint URL for your index management page.

          • credentialsSecretArn (string) –

            The Amazon Resource Name (ARN) of the secret that you created in Secrets Manager that is linked to your Pinecone API key.

          • fieldMapping (dict) –

            Contains the names of the fields to which to map information about the vector store.

            • metadataField (string) –

              The name of the field in which Amazon Bedrock stores metadata about the vector store.

            • textField (string) –

              The name of the field in which Amazon Bedrock stores the raw text from your data. The text is split according to the chunking strategy you choose.

          • namespace (string) –

            The namespace to be used to write new data to your database.

        • rdsConfiguration (dict) –

          Contains details about the storage configuration of the knowledge base in Amazon RDS. For more information, see Create a vector index in Amazon RDS.

          • credentialsSecretArn (string) –

            The Amazon Resource Name (ARN) of the secret that you created in Secrets Manager that is linked to your Amazon RDS database.

          • databaseName (string) –

            The name of your Amazon RDS database.

          • fieldMapping (dict) –

            Contains the names of the fields to which to map information about the vector store.

            • metadataField (string) –

              The name of the field in which Amazon Bedrock stores metadata about the vector store.

            • primaryKeyField (string) –

              The name of the field in which Amazon Bedrock stores the ID for each entry.

            • textField (string) –

              The name of the field in which Amazon Bedrock stores the raw text from your data. The text is split according to the chunking strategy you choose.

            • vectorField (string) –

              The name of the field in which Amazon Bedrock stores the vector embeddings for your data sources.

          • resourceArn (string) –

            The Amazon Resource Name (ARN) of the vector store.

          • tableName (string) –

            The name of the table in the database.

        • redisEnterpriseCloudConfiguration (dict) –

          Contains the storage configuration of the knowledge base in Redis Enterprise Cloud.

          • credentialsSecretArn (string) –

            The Amazon Resource Name (ARN) of the secret that you created in Secrets Manager that is linked to your Redis Enterprise Cloud database.

          • endpoint (string) –

            The endpoint URL of the Redis Enterprise Cloud database.

          • fieldMapping (dict) –

            Contains the names of the fields to which to map information about the vector store.

            • metadataField (string) –

              The name of the field in which Amazon Bedrock stores metadata about the vector store.

            • textField (string) –

              The name of the field in which Amazon Bedrock stores the raw text from your data. The text is split according to the chunking strategy you choose.

            • vectorField (string) –

              The name of the field in which Amazon Bedrock stores the vector embeddings for your data sources.

          • vectorIndexName (string) –

            The name of the vector index.

        • type (string) –

          The vector store service in which the knowledge base is stored.

      • updatedAt (datetime) –

        The time at which the knowledge base was last updated.

Exceptions