The first 500 people to use my link https://skl.sh/tinytechnicaltutorials03251 will get a 1-month free trial of Skillshare!
This video is sponsored by Skillshare.
In this hands-on AWS tutorial, I’ll walk you through how to use Amazon Textract, Comprehend, and Bedrock to extract, analyze, and interpret text from images and documents. We’ll start by creating a SageMaker Notebook, then run code against the different services to see how the results differ.
IMPORTANT: Be sure to stick around to the end where I’ll show you how to delete your resources.
🌟***LINKS REFERENCED IN THIS VIDEO***🌟
•GitHub repository with code for Textract, Comprehend and Bedrock: https://github.com/tinytechnicaltutorials/textract-comprehend-bedrock
•PRICING
o Textract: https://aws.amazon.com/textract/pricing/
o Comprehend: https://aws.amazon.com/comprehend/pricing
o Bedrock: https://aws.amazon.com/bedrock/pricing/
o SageMaker Notebook: https://aws.amazon.com/sagemaker-ai/pricing
•Amazon’s guidance on which services to use: https://aws.amazon.com/solutions/guidance/intelligent-document-processing-on-aws/
🌟***TIMESTAMPS***🌟
00:00 – AWS AI and ML services for doing optical character recognition (OCR) and intelligent document processing (IDP)
00:45 – What you need to follow along with the hands-on tutorial
02:08 – How much will this cost?
03:40 – Creating a new SageMaker notebook instance in SageMaker AI
05:00 – Adding permissions policies to the SageMaker IAM role for Textract, Comprehend and Bedrock
07:00 – Opening JupyterLab and creating the SageMaker notebook
07:35 – Retrieving code from the GitHub repository
08:03 – Using Textract to extract text from an image using Python
12:14 – Using Comprehend with Textract to pull out key phrases and sentiment
18:15 – Summarizing what we’ve done so far with Textract and Comprehend
18:46 – Using Amazon Bedrock with Anthropic’s Claude 3 Haiku to extract text and find meaning
21.48 – Amazon Textract vs. Comprehend vs. Bedrock – or should I use them together?
23:13 – IMPORTANT!! Delete your resources