n8n Conversational Sales Agent for WooCommerce with GPT-4, Stripe, and CRM Integration

This n8n workflow orchestrates a sophisticated conversational AI sales agent designed to interact with customers, answer product questions, process orders, and manage customer relationships for a WooCommerce store. It leverages advanced AI models, vector databases, and n8n's workflow capabilities to provide a seamless and intelligent customer experience.

What it does

This workflow simplifies and automates customer interactions for an e-commerce business by:

1. Triggering on Chat Messages: Initiates the conversation whenever a new chat message is received, acting as the entry point for customer inquiries.
2. Processing with an AI Agent: Routes the customer's message to an AI Agent, which utilizes various tools and language models to understand the query and formulate a response.
3. Leveraging OpenAI Chat Model: Employs an OpenAI Chat Model (likely GPT-4 or similar) for natural language understanding and generation, enabling human-like conversations.
4. Maintaining Conversation History: Utilizes a Simple Memory component to keep track of the conversation context, ensuring the AI agent remembers previous interactions.
5. Accessing Product Information (Vector Store): Queries a Qdrant Vector Store, which likely holds product descriptions, FAQs, and other relevant store data, to answer specific product-related questions. This is facilitated by an "Answer questions with a vector store" tool.
6. Loading and Splitting Documents: Uses a Default Data Loader and Token Splitter to prepare documents (e.g., product catalogs, policy documents) for embedding and storage in the vector database.
7. Generating Embeddings with OpenAI: Converts text data into numerical embeddings using OpenAI Embeddings, which are then stored in Qdrant for efficient semantic search.
8. Extracting Information: Employs an Information Extractor to parse structured data from customer messages, such as order details, customer preferences, or contact information.
9. Interacting with External Systems (via n8n Workflow Tools): Calls other n8n workflows as tools, enabling the AI agent to perform actions like:
   • Processing orders (e.g., with Stripe for payments, WooCommerce API).
   • Updating CRM records.
   • Retrieving specific customer or order details.
   • Sending notifications.
10. Making HTTP Requests: Executes generic HTTP requests, likely for interacting with WooCommerce API, Stripe API, or other external services not directly covered by dedicated n8n nodes.
11. Editing and Transforming Data: Uses a Set (Edit Fields) node to manipulate and format data as it flows through the workflow, ensuring compatibility between different nodes and services.
12. Integrating with Google Drive: Potentially stores or retrieves documents related to customer interactions or product information from Google Drive.
13. Supporting Google Gemini (Optional): Includes a Google Gemini Chat Model, suggesting flexibility to switch between OpenAI and Google's AI models.

Prerequisites/Requirements

To effectively use this workflow, you will need:

• n8n Instance: A running instance of n8n.
• OpenAI API Key: For the OpenAI Chat Model and Embeddings.
• Qdrant Instance: Access to a Qdrant vector database for storing and retrieving product information.
• WooCommerce Store: An active WooCommerce store with API access enabled.
• Stripe Account: For payment processing (if integrated into the linked n8n workflow tools).
• CRM System: Access to your CRM system (e.g., HubSpot, Salesforce, or a custom one) if CRM integration is part of the linked n8n workflow tools.
• Google Drive Account: If utilizing Google Drive for document storage.
• Google Gemini API Key (Optional): If you choose to use the Google Gemini Chat Model.
• Additional n8n Workflows: This workflow relies on other n8n workflows to act as "tools" for the AI agent (e.g., "process_order_workflow", "update_crm_workflow"). These sub-workflows need to be set up and configured separately.

Setup/Usage

1. Import the Workflow: Download the provided JSON and import it into your n8n instance.
2. Configure Credentials:
   • Set up your OpenAI API Key credentials in n8n.
   • Configure Qdrant credentials (host, API key if applicable).
   • Set up Google Drive credentials if you plan to use that node.
   • Configure Google Gemini credentials if you intend to use it.
3. Configure AI Agent:
   • In the "AI Agent" node, select your OpenAI Chat Model and Simple Memory.
   • Define the "Call n8n Workflow Tool" nodes, linking them to the specific n8n workflows you've created for actions like "process order," "update CRM," etc. Ensure these sub-workflows are properly configured to receive and return data.
   • Configure the "Vector Store Question Answer Tool" to use your Qdrant Vector Store and OpenAI Embeddings.
4. Populate Vector Store: Ensure your Qdrant vector store is populated with relevant product information, FAQs, and other knowledge base articles. This typically involves using the "Default Data Loader," "Token Splitter," and "Embeddings OpenAI" nodes in a separate workflow to ingest your data into Qdrant.
5. Configure Chat Trigger: Set up the "When chat message received" trigger according to your chat platform (e.g., a webhook for a custom chat interface, a dedicated chat integration).
6. Activate the Workflow: Once all configurations are complete, activate the workflow.

This workflow provides a robust framework for building an intelligent conversational sales agent. You will need to customize the "Call n8n Workflow Tool" nodes and their corresponding sub-workflows to match your specific WooCommerce, Stripe, and CRM integrations.