AWS Announces Amazon Comprehend for Intelligent Document Processing

Amazon Web Services announced earlier this month Amazon Comprehend for Intelligent Document Processing (IDP). This is a new natural language processing (NLP) service that uses machine learning (ML) to uncover relationships and connections in various document types.

Amazon Comprehend for IDP enables you to classify and extract entities from PDF documents, Microsoft Word files, and images directly without requiring you to extract the text first. Previously, businesses would typically need to pre-process those documents and convert them into some type of machine-readable text which can reduce the quality of the document context.

Some of the things that businesses can do with Amazon Comprehend for IDP include:

• Automating the extraction of insights from packets of legal briefs like contracts and court records.
• Securing documents by identifying and redacting Personally Identifiable Information (PII).
• Detecting customer sentiment and analyzing customer interactions.
• Automatically categorizing inbound support requests.
• Analyzing documents and indexing key phrases and entities.
• Classifying and extracting entities from financial services documents such as insurance claims.

How to get started with Amazon Comprehend for IDP

You can use Amazon Comprehend for IDP from the AWS Management Console, AWS SDKs, or the AWS Command Line Interface (CLI). Amazon Comprehend offers a free tier covering 50K units of text (5M characters) per API per month. Eligible APIs for the free tier include Key Phrase Extraction, Sentiment, Targeted Sentiment, Entity Recognition, Language Detection, Event Detection, Syntax Analysis, Detect PII, and Contains PII.

Beyond the free tier, Amazon Comprehend for IDP is priced per unit and per feature. Some sample pricing is shown in the table below.  

You can estimate your costs for Amazon Comprehend for IDP using the AWS Pricing Calculator. For regional availability, you can refer to the AWS Regional Services List.