⚒️ Decide multi‑agent vs simple workflows using Azure OpenAI GPT‑4o‑mini
⚡ 22 views · ⚒️ Engineering
Description
📘 Description
This workflow acts as an AI Multi-Agent Architecture Advisor for n8n. It receives a problem statement via webhook, uses Azure OpenAI (gpt-4o-mini) to decide whether the problem needs a multi-agent design or a simple workflow, then returns a styled HTML report showing the decision, recommended agents (if any), and the suggested step flow.
⚙️ Step-by-Step Flow
- Receive Problem Description via POST (Webhook)
- Accepts a POST payload containing a description field.
- Extract Request Body (Code)
- Strips the wrapper and outputs only the body JSON to simplify downstream prompts.
- Multi-Agent Architecture Decision Agent (AI Agent)
- Analyzes the problem description and outputs one of:
- Decision: Multi-Agent Required + agent definitions + workflow steps
- Decision: Not Required + minimal simplified workflow nodes
- Parse Decision, Agents & Steps (Code)
- Extracts three structured fields from the
AI output:
decision
- agents[] (name, purpose, node, reason)
- workflow_flow[] (ordered steps)
- Build HTML Architecture Report (Code)
- Renders a clean card-based HTML dashboard:
- Decision badge (multi vs simple)
- Agent cards (if present)
- Workflow flow chips (steps)
- Return HTML Report to Caller (Respond to Webhook)
- Returns the HTML report directly as the webhook response.
🧩 Prerequisites
- • Azure OpenAI credential with an active gpt-4o-mini deployment
- • n8n webhook endpoint exposed to the caller (or via tunnel)
💡 Key Benefits
✔ Fast “multi-agent vs simple” decisioning ✔ Outputs an actionable architecture, not generic advice ✔ HTML report is ready to embed in internal tools (Base44/UI) ✔ Structured parsing makes it easy to store or extend later
👥 Perfect For
- Automation agencies doing solution design calls
- Teams standardizing how they choose agent-based workflows
- Internal tooling that needs instant architecture recommendations
🔗 Nodes Used
Webhook, AI Agent, Azure OpenAI Chat Model
📥 Import
Download workflow.json and import into n8n:
Workflow menu → Import from File