Build an AI Application for Document Classification with Orkes Conductor
In today's data driven organizations, the classification of documents is critical, and doing so inefficiently can be time consuming, resource intensive, and error prone. Jeff Bull (Developer Community Manager, Orkes) shows you how to build an AI-powered document classification app using Orkes Conductor. This workflow will identify and categorize PDF documents (W-2 forms, driver’s licenses, pay stubs, etc.) while handling errors for unsupported file types.
🔗 Resources:
📖 Blog with full workflow JSON & instructions: https://orkes.io/blog/ai-app-for-docu...
🖥 Try Orkes Developer Edition for free: https://developer.orkescloud.com
💬 Join our upcoming events: https://orkes.io/events
01:00 - Workflow Breakdown: Switch Tasks, OCR & LLM Integration
02:00 - Creating the PDF Text Extraction Task
03:00 - Building the Full Workflow & Using Predefined JSON
04:00 - Setting Up an OCR Worker Task (Node.js)
05:00 - Configuring Access Keys & External Worker Registration
06:00 - Integrating LLM for Document Classification
08:00 - Testing the LLM Prompt & Model Selection
11:00 - Running the Workflow with Sample Inputs
13:00 - Handling Errors & Unsupported File Types
14:00 - Advanced Features & What’s Next
Join the Conductor community on:
Slack - https://orkes-conductor.slack.com
Discourse - https://community.orkes.io