ad
ad

Getting Started with Intelligent Document Processing: Part 1

Science & Technology

Getting Started with Intelligent Document Processing: Part 1

Introduction

Hello everyone and welcome to Mule Bites! In this video, we are going to get started with Intelligent Document Processing, also known as IDP. I'm Akh Savant, Senior Developer Advocate at Salesforce, and today we are going to see how to harness the power of IDP to automate your document processing. So, let's get started!

What is IDP?

MuleSoft's IDP is a powerful tool that helps you extract and analyze information from your documents using AI capabilities. Under the hood, it uses powerful technologies like AWS Textract to extract information from your documents. Additionally, you can further integrate it with Anypoint Platform, Flow, and RPA. Today, I'll show you how to design your first document action.

Access Rights and Setup

Before getting started with IDP, you need to ensure that you have valid access rights to use it. Let's walk through how to set it up:

  1. Access Management:

    • Go into Access Management on the Anypoint Platform.
    • Select your user and add permissions. Type in "document actions" and select all three permissions available.
    • Choose your business group and add all permissions to your account.

  2. Navigating to IDP:

    • Head over to Intelligent Document Processing under the Automation tab.
    • This will bring you to the landing page of IDP, where you can see any pre-existing document actions and create new ones.

Creating Your First Document Action

Next, let's create our first document action:

  1. Choose Template:

    • Select a new document action and you'll be presented with different templates like Invoice, Purchase Order, and a generic template.
    • For this tutorial, we'll use the Purchase Order template.
    • Name the document action (e.g., "Standard Purchase Order") and add a suitable description.

  2. Uploading and Extracting Information:

    • Upload a file to extract information from. This can be in formats like PDF or JPEG.
    • Click Run to start the extraction process.
    • The technology behind the scenes, AWS Textract, will extract the information, and it will display in the fields tab.

  3. Configuring Fields:

    • Review the extracted fields and mark those that were not found as optional or exclude them.
    • You can click on a specific field (e.g., Tax) to set the confidence threshold, and label the field as required or optional.

  4. Handling Tables and Prompts:

    • If your document contains tables, the tool will automatically extract and display the relevant information.
    • Add new prompts if certain key fields like PAN number or GST number were not extracted.
    • Specify the questions for these prompts and set confidence levels.

  5. Assigning a Reviewer:

    • Click on the add button to assign a reviewer.
    • Save and publish the document action to Anypoint Exchange.

Conclusion

We have now seen how to build a simple document action in IDP. In upcoming sessions, we'll learn how to integrate document action with Anypoint Platform, Flow, and RPA. Below are some resources that might be helpful for you to get started with IDP.

Additional Resources

  • MuleSoft Documentation
  • Anypoint Platform Overview
  • AWS Textract Documentation

If you liked this video, don't forget to like and subscribe to our channel, and check out other videos from the Mule Bites playlist. Thank you for watching!

Keywords

  • Mule Bites
  • Intelligent Document Processing (IDP)
  • AI Capabilities
  • AWS Textract
  • Anypoint Platform
  • Document Action
  • Automation
  • Purchase Order Template
  • Confidence Threshold
  • Reviewer Assignment

FAQ

  1. What is Intelligent Document Processing (IDP)?

    • IDP is a tool that automates document extraction and analysis using AI and technologies like AWS Textract.

  2. What formats are supported for document uploads in IDP?

    • IDP supports multiple formats, including PDF and JPEG.

  3. What preliminary setup is needed before using IDP?

    • You must ensure you have the necessary access rights in Access Management on the Anypoint Platform.

  4. What templates are available in IDP?

    • As of now, templates available include Invoice, Purchase Order, and a generic template.

  5. How do you set a confidence threshold for extracted fields?

    • You can set the confidence threshold by clicking on a specific field in the fields tab and adjusting the settings.

  6. Can you add custom prompts in IDP?

    • Yes, you can add up to 30 custom prompts for your document action.

  7. How do you assign a reviewer for a document action?

    • You can assign a reviewer by clicking the add button in the reviewer section and saving the settings.

  8. Where can I find more resources to get started with IDP?

    • MuleSoft Documentation and the Anypoint Platform Overview are ideal places to start.