Hi all,
I'm building an AI Agent for an HR Welfare Claim automation and need guidance on the Smart Document AI Skill.
Use case:
When an HR Case with Category = "Welfare Claim" is created, employees attach several supporting documents. One HR Case can contain multiple individual claim requests. I need an AI Agent that reads all attachments, classifies & extracts each document, and populates a custom child table (`Welfare Claim Recap`) linked to the HR Case.
Already tested: Document Q&A in Now Assist chatbot - uploaded a document, the AI read it accurately. Now I want to operationalize this via an AI Agent.
Questions:
1. How do I add the Smart Document skill to a custom AI Agent in AI Agent Studio?
2. How does the agent access attachments from the triggering HR Case — passed automatically as context, or do I need to wire it explicitly?
3. What's the recommended pattern for multiple attachments of different types on one case (loop, batch, or chain skills)?
4. How does Smart Documents skill relate to Now Assist in Document Intelligence use cases — which one fits my scenario?
Any reference articles, demo flows, or implementation experience would be hugely appreciated. Thanks!
