Text classification or multi AI Agent Tool, Which approach for scalable AI Orchestration?

[Phat_Truong]

I’m debating between two ways to route tasks in my n8n AI workflows. I’d love to hear your experience on:

1. Text Classification + Switch Node: Using an LLM to categorize intent first, then routing via standard nodes.

2. Multi-AI Agent (Agent as a Tool): A Master Agent that calls sub-agents as tools
   I’d love to hear your experiences and “war stories” on which one works better