Amazon Augmented AI (A2I) is a service that enables the incorporation of human reviews into AI workflows, ensuring greater accuracy and reliability in machine learning predictions. When paired with Amazon Comprehend, an NLP service for analyzing text, it allows for building robust sentiment analysis systems that can automatically process customer feedback and involve customized reviewer responses for flagged negative sentiments.
In this Cloud Lab, you will set up an Amazon S3 bucket to store input files containing customer feedback and output files with processed results. You will configure IAM roles to securely enable access for Amazon A2I and Lambda functions. The lab will guide you through creating an A2I human review workflow by designing a custom UI template for reviewers and building a private review team to assess flagged feedback. You will also develop a Lambda function to automate the process, where customer feedback is analyzed using Amazon Comprehend, and negative sentiments are flagged. The S3 bucket stores reviewer responses written to address the flagged negative feedback.
After completing this Cloud Lab, you will have the skills to build a human-in-the-loop sentiment analysis workflow using Amazon A2I and Comprehend. You’ll also learn how to automate sentiment analysis processes, enhance customer interactions, and improve the quality of CRM systems with human-augmented AI workflows.
A high-level architecture diagram for this Cloud Lab is given below:
