Route WhatsApp Intents with LiveConnect & GPT-4o-mini in n8n

Route WhatsApp intents to the right team in n8n—automatically

This n8n workflow receives LiveConnect Flowbot callback messages from WhatsApp, uses GPT-4o-mini with conversation memory to classify the user’s intent, drafts a short Spanish reply, and then routes the conversation to the correct LiveConnect callback response (Sales, Support, Agenda, or General).

What this workflow does

• Receives user messages via LiveConnect callback webhook from your Flowbot, including a session ID and message text, protected with a shared secret.
• Classifies intent and drafts a reply using OpenAI (GPT-4o-mini) with a per-session conversation memory. The model outputs structured JSON in the format { intencion, mensaje }.
• Parses the structured output to extract intencion and the drafted Spanish message.
• Routes automatically by intent:
  • Sales → sends the drafted text back via LiveConnect and saves an intent variable for Sales.
  • Support → sends the drafted transfer message and delegates to a selected LiveConnect team for human assistance.
  • Agenda → sends the drafted scheduling message, saves the Agenda intent, and prompts for a preferred day and time.
  • General → uses a fallback path for “other” questions.

Use cases

• A prospect asks about pricing: route to Sales and respond instantly.
• A user needs help troubleshooting: route to Support and transfer to a human team.
• A customer wants an appointment: route to Agenda and collect preferred day/time.
• General questions: answer via the General callback response.

Technical details (n8n stack)

• LiveConnect Callback Trigger and LiveConnect callback response nodes
• OpenAI Chat Model: gpt-4o-mini
• Intent routing with switch
• Structured output handling with n8n-nodes-langchainoutput (Structured JSON Parser)
• Conversation memory using n8n-nodes-langchainmemory buffer window
• Agent/LLM orchestration with n8nn8n-nodes-langchainagent and chat integration nodes
• Optional workflow guidance using a sticky note

Set up requirements: add your OpenAI credential and confirm gpt-4o-mini, add LiveConnect API credentials, and configure a secure shared secret in the LiveConnect Callback Trigger.