Hi everyone,
I’m exploring the new Agentic AI capabilities introduced in the ServiceNow Yokohama release, specifically how to integrate AI agents built in AI Agent Studio with Virtual Agent conversations.
Here’s what I’m trying to do:
Build an AI agent in AI Agent Studio to handle a specific task (e.g., record lookup, case creation, or workflow orchestration).
Expose this agent so that Virtual Agent can call it dynamically during a chat, based on user input.
Pass required inputs (like record IDs or user queries) from the Virtual Agent topic to the AI agent, and handle the outputs back in the conversation flow.
Questions:
✅ What’s the recommended approach in Yokohama to link or call an AI agent from Virtual Agent Designer?
✅ Once the AI agent is published, how do I:
Access it as an action node inside Virtual Agent Designer?
Map the input/output variables correctly?
✅ Are there new best practices or design patterns to orchestrate multiple AI agents within a single conversation?
✅ Any considerations around testing, governance, or monitoring when using Agentic AI with Virtual Agent?
I’d love to hear:
How others have approached this integration
Tips, lessons learned, or official docs you found especially useful
Thanks in advance for sharing your insights!
