Welcome to the documentation for AWS Textract. Textract is a fully managed machine learning service that automatically extracts printed text, handwriting, and other data from scanned documents or photos. This documentation will guide you through the process of getting started with Textract and provide detailed information on its features and functionality.

Getting Started

To get started with AWS Textract, you need an AWS account. If you don’t have one, you can create a new AWS account at Once you have an account, follow these steps to get started:

  • Create or select an AWS identity and access management (IAM) role to grant Textract access to other AWS resources on your behalf. This IAM role will be used to provide access to your S3 buckets and publish Amazon SNS notifications.
  • Set up an Amazon S3 bucket to store the input documents and the output files. Textract will use this bucket to store extracted data and any additional information related to your documents.
  • Create a new AWS Lambda function to process the extracted data. The Lambda function will enable you to perform custom logic on the extracted data and integrate Textract into your existing workflows.
  • Configure the Amazon Simple Notification Service (SNS) to receive notifications when Textract processes your documents.


AWS Textract offers several powerful features to help you extract data from your documents effectively. Here are some of the key features:

  • Optical Character Recognition (OCR): Textract uses OCR to detect and extract printed text from images and scanned documents.
  • Handwriting Recognition: Textract can also recognize and extract text written by hand.
  • Structured Data Extraction: With the help of machine learning models, Textract can identify and extract key-value pairs, tables, and other structured data from your documents.
  • Form Extraction: Textract can automatically detect and extract data from forms, including checkboxes, radio buttons, and text fields.
  • Entity Extraction: Textract can identify and extract entities such as names, addresses, phone numbers, and dates from your documents.

Tips and Best Practices

To make the most out of AWS Textract, consider these tips and best practices:

  • Ensure that your input documents are of high quality and resolution to improve the accuracy of the extraction process.
  • Use the appropriate block type mapping to extract different types of data accurately. For example, use the KEY_VALUE_SET mapping to extract key-value pairs and the TABLE mapping to extract tabular data.
  • Regularly evaluate and fine-tune your machine learning models to improve extraction accuracy.
  • Take advantage of the Textract APIs to integrate Textract into your applications and workflows seamlessly.

Additional Resources

For more information on working with AWS Textract, refer to the following resources: