Streamline document processing with DocAI, Pega’s intelligent automation that requires no training or complex prompts. Powered by Pega GenAI, it effortlessly classifies and extracts data from diverse document types, extending the capabilities of the Pega Platform
DocAI can be used to classify documents and then extract information according to the classification. It supports data extraction from tables as well as checkboxes and works with printed multi-page documents as well as handwritten documents. The solution supports English and most Western European languages and ensures that the Large Language Model (LLM) providers do not store or use any data uploaded to DocAI.
Demo**:**
https://players.brightcove.net/1519050010001/default_default/index.html?videoId=6385979140112
Steps:
- Add an attachment field for “ClaimsAttachment” in your data model, and configure the attachment field in the workflow for users to upload their Claim document.
- In App Studio → AI Designer → GenAIConnect and create new.
- Provide the Class, Name and description of the Gen AI connect.
- AutoPilot generates the necessary artifacts based on the description provided for GenAIConnect rule.
- Set up GenAI Connect in your workflow, specifying request and response parameters to ensure extracted data maps to your business rules.
- On the Advanced tab of the GenAI rule, enable attachments for analysis.
- Go to Case type → Add an Automated step (Gen AI connect)
- Design the UI to display Claim Information as part of the case workflow.
- In this example, I’ve mapped the response to Unstructured, so creating a data transform to parse the json response to Clipboard structure using the below activity.
@pxConvertStringToPage(tools,.PlanMemberInformation,.ExtractClaimData,“json”)
Reference:
Pega GenAI Series
Enjoyed this article? See more similair articles in Pega Cookbook Recipes 
series.
