Customer Support & Lead Collection Chatbot with RAG (GPT-4o, Pinecone)

This n8n workflow automates a sophisticated customer support and lead collection chatbot using Retrieval Augmented Generation (RAG) with GPT-4o and Pinecone, designed to provide intelligent responses and gather user information.

What it does

This workflow sets up a conversational AI agent that can interact with users, answer questions using a knowledge base (Pinecone), and maintain a conversational history.

1. Listens for Chat Messages: The workflow is triggered whenever a new chat message is received from a user.
2. Manages Conversational Memory: It maintains a short-term memory of the conversation to provide context for AI responses.
3. Utilizes a Vector Store for RAG: It connects to a Pinecone vector store, enabling the AI to retrieve relevant information from a predefined knowledge base to answer user questions accurately (RAG).
4. Generates Responses with OpenAI Chat Model: It uses an OpenAI Chat Model (e.g., GPT-4o) to process user input, incorporate context from memory and the vector store, and generate natural language responses.
5. Embeds Data for Vector Store: It uses OpenAI Embeddings to convert text into numerical vectors, which are then used by the Pinecone vector store for efficient similarity search.
6. Acts as an AI Agent: The core of the workflow is an AI Agent that orchestrates the interaction between the chat model, memory, and vector store tool to provide comprehensive and contextually aware responses.

Prerequisites/Requirements

To use this workflow, you will need:

• n8n Instance: A running n8n instance.
• OpenAI API Key: For the OpenAI Chat Model and Embeddings.
• Pinecone Account: With an existing index for your knowledge base.
• Langchain Credentials: Configured within n8n for OpenAI and Pinecone.
• A Chat Service: (e.g., Telegram, Slack, Discord) configured with the Chat Trigger node to receive messages.

Setup/Usage

1. Import the Workflow: Download the provided JSON and import it into your n8n instance.
2. Configure Credentials:
   • OpenAI Chat Model: Configure your OpenAI API Key credentials.
   • Embeddings OpenAI: Configure your OpenAI API Key credentials.
   • Pinecone Vector Store: Configure your Pinecone API Key and environment credentials. Ensure your Pinecone index is populated with your knowledge base data.
3. Configure Chat Trigger: Set up the "When chat message received" node to connect to your desired chat service (e.g., Telegram Bot, Slack App).
4. Activate the Workflow: Save and activate the workflow.

The chatbot will now be ready to receive messages from your configured chat service, provide answers from its knowledge base, and maintain conversational context.