Inbound leads go cold while you're busy. Here's a 5-node n8n workflow that scores every form submission with an LLM and pings Slack only for the leads worth your time.
The flow
Webhook → Config → Build Prompt → LLM (HTTP) → Parse → IF score ≥ 70 → Slack → Respond
I don't use a vendor-specific AI node. The LLM call is a plain HTTP Request to /chat/completions, so it works with OpenAI, Groq, OpenRouter, or local Ollama. Just change baseUrl.
The prompt that makes it reliable
You are a B2B sales analyst. Score inbound leads 0-100 for fit and intent using BANT.
Return JSON: {score, tier, summary, reasons[], next_step}
Add response_format: {type: "json_object"} and the IF node can branch on {{$json.result.score}} without breaking.
Try it
curl -X POST http://localhost:5678/webhook-test/lead-qualifier -H 'Content-Type: application/json' \
-d '{"name":"Priya","company":"Acme SaaS","message":"Need an onboarding bot, budget $2k, this month"}'
Response: {"score":82,"tier":"hot","next_step":"Book a call this week..."}, and a Slack ping.
Get the workflow
The JSON is free on GitHub: https://github.com/arkintof/n8n-ai-agent-workflows
Want more like it (support agent, invoice/GST extractor, competitor digest, resume screener)? The repo README lists them: https://github.com/arkintof/n8n-ai-agent-workflows
Written with AI assistance; workflow built and tested by me.
Top comments (0)