Building a Data Classification Agent

Learn how to create an AI-powered document classification system with Prisme.ai

This tutorial guides you through building an intelligent document classification system using Prisme.ai. You’ll create a solution that automatically categorizes uploaded documents into types such as invoices, CVs, quotes, contracts, and more using the power of generative AI.

What You’ll Build

A complete document management system with:

• A user-friendly document upload interface
• Automatic AI-based document classification
• A searchable and filterable document repository
• Document management capabilities (view, categorize, delete)

This solution integrates Knowledges for classification with Collection for document storage, creating a powerful yet simple document management system.

Prerequisites

Before starting this tutorial, make sure you have:

• An active Prisme.ai account
• The Knowledge Client app installed in your workspace
• The Collection app installed in your workspace
• Basic familiarity with Prisme.ai Builder

Step 1: Creating Your Document Management Workspace

Let’s start by setting up a dedicated workspace for our document management system:

1. Access Builder

   Log in to your Prisme.ai account and navigate to the Builder product

2. Create a New Workspace

   Click the “Create Workspace” button to start a new project

3. Configure Workspace Settings

   Name your workspace “Document Management System” (or a name of your choice) and configure any additional settings as needed

Step 2: Building the Document Upload Interface

Now, let’s create a user-friendly interface for document uploads and management:

1. Create a New Page

   In your workspace, navigate to the Pages section and click “Create Page”

2. Configure Page Properties

   • Name your page “Document Upload Form”
   • Set the slug to “upload-docs” (this will be used in the URL)
3. Add a Form Block

   Add a Form block to your page for document uploads with the following configuration:

   slug: upload-docs
   name: Document Upload Form
   blocks:
     - slug: Form
       schema:
         type: object
         properties:
           attachment:
             type: string
             ui:widget: upload
             title: Document Upload
             description: Attach your document here.
             ui:options:
               upload:
                 accept: .pdf
           description:
             type: textarea
             title: Document Description
             placeholder: Describe the document here.
       onSubmit: save
   
4. Add a Data Table Block

   Add a DataTable block to display uploaded documents and their classifications:

   - slug: DataTable
     columns:
       - label: ID
         key: id
         type: string
       - label: Description
         key: description
         type: string
       - label: Category
         key: category
         type: string
       - label: Actions
         actions:
           - label: Delete
             action:
               type: event
               value: deleteData
               payload:
                 id: ${_id}
     config:
       customProps:
         loading: true
       title:
         fr: Uploaded Documents
         en: Uploaded Documents
     updateOn: updateData
     onInit: initData
     data:
       - {}
   
5. Save Your Page

   Save your page configuration. You should now have a form for uploads and a table to display documents

Step 3: Creating the Document Classification Automation

Next, let’s set up the automation that will classify documents when they’re uploaded:

1. Navigate to Automations

   In your workspace, go to the Automations section

2. Create Document Classification Automation

   Create a new automation named “Manage Form” with the following configuration:

   slug: manage-form
   name: Manage Form
   do:
     - Knowledge Client.chat-completion:
         messages:
           - role: system
             content: 'Classify the document within these categories: invoice, CV, quote, contract, others. Provide your categorization based on the content.'
           - role: user
             content: '{{payload.description}}'
         output: data
     - conditions:
         '{{data.response}}':
           - set:
               name: payload.category
               value: '{{data.response}}'
               type: merge
         default:
           - emit:
               target:
                 currentSocket: true
               options:
                 persist: true
               event: error
               payload: '{{data}}'
           - set:
               name: payload.error
               value: '{{data.error.message}}'
               type: merge
     - Collection.insert:
         data: '{{payload}}'
     - emit:
         target:
           currentSocket: true
         options:
           persist: false
         event: initData
   when:
     events:
       - save
   output: '{{payload}}'
   
3. Create Data Retrieval Automation

   Create another automation named “Get Data” to fetch and display documents:

   slug: get-data
   name: Get Data
   do:
     - Collection.find:
         query: {}
         output: data      
         sort:
           createdAt: -1        
     - emit:
         event: updateData
         payload:
           data: '{{data}}'
           customProps:
             loading: false  
   when:
     events:
       - initData   
   output: ''
   
4. Create Document Deletion Automation

   Create a final automation named “Delete Data” to handle document removal:

   slug: delete-data
   name: Delete Data
   do:
     - conditions:
         '!{{payload.data._id}}':
           - break: {}            
         default: []
     - Collection.deleteOne:
         query:
           _id: '{{payload.data._id}}'
     - emit:
         target:
           currentSocket: true
         options:
           persist: true
         event: initData   
   when:
     events:
       - deleteData
   output: ''
   
5. Deploy Your Automations

   Save all automations and make sure they’re properly configured to work together

Step 4: Testing and Using Your Document Classification System

Now it’s time to test your document management system:

1. Access Your Document Upload Page

   Navigate to your document upload page using the URL format: workspace-slug.pages.host/lang/upload-docs

2. Upload a Test Document

   • Upload a sample document (PDF) through the form
   • Add a description that gives context about the document’s content
   • Submit the form
3. Verify Classification

   After a moment, the document should appear in the data table with its AI-determined classification

4. Test Document Management

   Try deleting a document to ensure the delete functionality works correctly

Step 5: Version Control and Deployment

To finalize your document classification system:

1. Push Your Changes

   Use the “Push” button in your workspace to save the current state of your project

2. Create a Version

   Create a version of your workspace for easy rollback if needed in the future

3. Configure Access Controls

   Set up appropriate Role-Based Access Control (RBAC) to determine who can use your document management system

Monitoring and Improving Your System

After deployment, regularly check the system’s performance:

1. Monitor Activity Logs

   Review the activity logs to track document uploads and classifications

2. Review Classification Accuracy

   Periodically check if documents are being classified correctly and refine your system as needed

3. Optimize for Performance

   Make adjustments to improve speed and accuracy as your document volume grows

Extending Your Document Classification System

Your base system is powerful, but consider these enhancements:

• Advanced Classification: Train a more specialized model for your specific document types
• Content Extraction: Extract key information from documents based on their category
• Automated Workflows: Trigger specific actions based on document type (e.g., route invoices to accounting)
• Search Functionality: Add search capabilities to find documents by content or metadata
• Integration: Connect with other systems like CRM or ERP platforms

Next Steps

Build a Webhook Handler

Learn how to create webhook integrations for your document management system

Explore AI Contact Routing

Discover how to route documents to the right department automatically

Create a RAG Agent

Build an agent that can answer questions about your document repository

Integrate with Websites

Connect your document system to web content for enhanced capabilities