n8n Multi-LLM Customer Support Chatbot for WordPress

This n8n workflow provides a flexible and powerful customer support chatbot solution, designed to integrate with WordPress and leverage multiple Large Language Models (LLMs) for intelligent responses. It acts as a central hub, receiving user queries (likely from a WordPress site via a webhook), processing them with an AI agent, and then responding appropriately.

What it does

This workflow simplifies and automates the process of handling customer support queries by:

1. Receiving incoming requests: It listens for incoming HTTP requests via a webhook, which serves as the entry point for customer queries (e.g., from a WordPress form or custom integration).
2. Initializing an AI Agent: It sets up an advanced AI Agent (powered by LangChain) to process the incoming query.
3. Configuring AI Memory: It utilizes a "Simple Memory" (Buffer Window Memory) to maintain conversational context, allowing the chatbot to remember previous interactions within a session.
4. Selecting an LLM: It provides a selection of multiple LLMs (OpenAI, Anthropic, Google Gemini, OpenRouter, xAI Grok) as the core intelligence for the AI Agent, allowing for flexibility and choice based on performance, cost, or specific model capabilities.
5. Enabling "Think" Tool: It includes a "Think" tool for the AI Agent, which can be used for internal reasoning or complex logic before generating a response.
6. Responding to the webhook: After the AI Agent processes the query and generates a response, the workflow sends this response back to the original webhook caller.
7. Conditional Logic (Placeholder): It includes an "If" node, suggesting potential for conditional routing or processing based on the incoming data or AI response (though currently not connected in the provided JSON).
8. Data Transformation (Placeholder): An "Edit Fields (Set)" node is present, indicating the capability to transform or manipulate data at various stages of the workflow (though currently not connected).

Prerequisites/Requirements

To use this workflow, you will need:

• n8n Instance: A running n8n instance (self-hosted or cloud).
• Webhook Integration: A system (e.g., WordPress plugin, custom code) capable of sending HTTP POST requests to the n8n webhook URL.
• API Keys for desired LLMs:
  • OpenAI API Key (for OpenAI Chat Model)
  • Anthropic API Key (for Anthropic Chat Model)
  • Google Gemini API Key (for Google Gemini Chat Model)
  • OpenRouter API Key (for OpenRouter Chat Model)
  • xAI Grok API Key (for xAI Grok Chat Model)
• LangChain Nodes: Ensure the @n8n/n8n-nodes-langchain package is installed in your n8n instance.

Setup/Usage

1. Import the Workflow:
   • Download the provided JSON file.
   • In your n8n instance, click on "Workflows" in the left sidebar.
   • Click "New" -> "Import from JSON" and paste the workflow JSON or upload the file.
2. Configure Webhook Trigger:
   • Locate the "Webhook" node (ID: 47).
   • Copy the "Webhook URL" displayed in the node settings. This URL will be used by your WordPress site or other integration to send queries.
3. Configure LLM Credentials:
   • Choose your preferred LLM node (e.g., "OpenAI Chat Model", "Anthropic Chat Model").
   • Click on the node and select or create a new credential for the respective service.
   • Enter your API key for the chosen LLM.
4. Configure AI Agent:
   • Click on the "AI Agent" node (ID: 1119).
   • Ensure the "Language Model" and "Memory" inputs are correctly connected to your chosen LLM and the "Simple Memory" node.
5. Activate the Workflow:
   • Once all configurations are complete, click the "Activate" toggle in the top right corner of the n8n editor to enable the workflow.
6. Integrate with WordPress:
   • Set up your WordPress site (e.g., using a form plugin or custom code) to send user queries as HTTP POST requests to the n8n Webhook URL. The request body should contain the user's message in a format the workflow expects (you may need to adjust the "AI Agent" node's input if the incoming data structure differs).
7. Test: Send a test query from your WordPress site to verify the chatbot responds correctly.

Note: The "If" and "Edit Fields" nodes are currently disconnected in the provided JSON. To utilize their functionality, you would need to connect them into the flow and configure their conditions/transformations as per your requirements.