Answer: Amazon Textract + Amazon Bedrock (Titan Text or Claude)

## Explanation **Amazon Textract + Amazon Bedrock (Titan Text or Claude)** is the correct combination because: 1. **Amazon Textract** is an AWS service specifically designed for extracting text and data from scanned documents, PDFs, and images. Clinical-trial documents often come in PDF format, and Textract can accurately extract the text content from these documents. 2. **Amazon Bedrock** is a fully managed service that provides access to foundation models (FMs) from leading AI companies, including Amazon's Titan models and Anthropic's Claude models. These models are excellent for natural language processing tasks such as: - **Summarization**: Creating concise summaries of lengthy clinical-trial documents - **Information extraction**: Identifying and extracting key findings, results, and important data points - **Content analysis**: Understanding complex medical terminology and research findings **Why the other options are incorrect:** - **B) Amazon Rekognition + Amazon Translate**: Rekognition is for image and video analysis (computer vision), not text document processing. Translate is for language translation, not summarization or key finding extraction. - **C) Amazon Transcribe + Amazon S3**: Transcribe is for converting speech to text (audio processing), not for processing text documents. S3 is just storage and doesn't provide document analysis capabilities. - **D) Amazon Polly + Amazon Comprehend**: Polly is for text-to-speech conversion, which is the opposite of what's needed. Comprehend is for natural language processing but is better suited for general text analysis rather than the specialized document extraction and summarization needed for clinical-trial documents. The Textract + Bedrock combination provides a complete solution: Textract extracts the text from documents, and Bedrock's advanced language models process that text to create summaries and extract key findings.

Author: Ritesh Yadav