Foundation Model Integration, Data Management, and Compliance II
Explore how to apply data validation and compliance checks when integrating foundation models within AWS Generative AI workflows. Understand best practices for preprocessing, governance, and retrieval strategies to ensure secure, accurate document handling and high-quality model inputs across various use cases and industries.
We'll cover the following...
Question 6
A company is building a document processing pipeline that ingests PDF files from Amazon S3 before sending extracted text to an Amazon Bedrock foundation model for summarization. The company wants to ensure data quality standards such as non-empty content, minimum text length, and valid metadata before the documents are consumed by the foundation model.
Which solution is the most appropriate for implementing these validation checks?
A. Use AWS Glue Data Quality rules to validate extracted text and metadata before passing the data to the FM.
B. Invoke the Amazon Bedrock model and rely on prompt instructions to reject invalid documents.
C. Use Amazon Comprehend to detect entities and infer whether ...