AWS::Bedrock::DataSource CustomTransformationConfiguration

Settings for customizing steps in the data source content ingestion pipeline.

You can configure the data source to process documents with a Lambda function after they are parsed and converted into chunks. When you add a post-chunking transformation, the service stores chunked documents in an S3 bucket and invokes a Lambda function to process them.

To process chunked documents with a Lambda function, define an S3 bucket path for input and output objects, and a transformation that specifies the Lambda function to invoke. You can use the Lambda function to customize how chunks are split, and the metadata for each chunk.

Syntax

To declare this entity in your AWS CloudFormation template, use the following syntax:

{
  "IntermediateStorage" : IntermediateStorage,
  "Transformations" : [ Transformation, ... ]
}

  IntermediateStorage: 
    IntermediateStorage
  Transformations: 
    - Transformation

Properties