A hospital deployed a RESTful API using Amazon API Gateway and AWS Lambda to upload reports in PDF and JPEG formats. The hospital needs to modify the Lambda code to identify protected health information (PHI) in the reports. Which solution meets the requirement with the LEAST operational overhead?

The correct solution uses Amazon Textract for text extraction and Amazon Comprehend Medical for PHI identification. Textract is a fully managed machine learning service that automatically extracts text, handwriting, and data from scanned documents, including PDFs and images (like JPEGs), making it ideal for the report formats. Comprehend Medical is a specialized natural language processing (NLP) service that uses machine learning to extract and identify medical information, including PHI, from unstructured clinical text. This combination offers the least operational overhead because both are fully managed services, requiring no machine learning expertise or infrastructure management from the hospital. Using existing Python libraries would involve significant development effort and ongoing maintenance for both text extraction and PHI identification, increasing operational overhead. Amazon SageMaker is a powerful machine learning service but requires building, training, and deploying custom models, which is more complex and has higher operational overhead than using a pre-trained service like Comprehend Medical. Amazon Rekognition is primarily for image and video analysis, including text detection, but Textract is specifically designed for document text extraction and provides more robust capabilities for this use case.