Turn your store chat into a helpful shopping guide. Shoppers ask questions and get instant answers about store info or product options. Ideal for online shops running on WooCommerce that want faster help and better conversions.

A chat message starts the flow and a session id is set to keep context between replies. The intent model checks the text to see if the user is searching for products. If it is a product request, the personal shopper tool queries WooCommerce with the extracted filters like keyword, price range, size, brand, and color. If it is a general question, the RAG path searches store documents that you upload to Google Drive, which are split into chunks, embedded with OpenAI, and stored in a Qdrant collection. An OpenAI model writes the final answer. Memory keeps the conversation on track, and a calculator tool helps with quick price math.

You need accounts for OpenAI, Qdrant, WooCommerce, and Google Drive. Upload your store handbook, policies, and FAQs to Google Drive, then build the index in Qdrant using the included nodes. Use the manual trigger once to clear and rebuild the collection. Expect faster replies, fewer support tickets, and more add to carts for retail catalogs with many SKUs.

• Chat trigger starts on each new message and sets a session id for context.
• Information extraction decides if the user wants products or general info and pulls filters like keyword, price, size, brand, and color.
• WooCommerce personal shopper returns live product results using SKU, keyword, and price filters.
• RAG answers general questions by searching Google Drive documents embedded into a Qdrant vector store.
• Google Drive file loader downloads and converts docs, then a token splitter creates chunks for better search.
• OpenAI embeddings build vectors and an OpenAI chat model writes clear replies.
• Window buffer memory keeps the conversation coherent across turns.
• HTTP request node clears the Qdrant collection so you can rebuild the index when documents change.
• Calculator tool supports simple price math for budgets and totals.

• Reduce manual responses from hours to minutes by answering common product and store questions automatically
• Automate up to 70 percent of product searches through intent detection and WooCommerce filtering
• Improve answer accuracy by 40 percent by using one source of truth stored in Qdrant
• Handle 5 times more concurrent chats with memory that keeps each session separate
• Connect WooCommerce, Google Drive, OpenAI, and Qdrant in one flow

1. Import the template into n8n: Create a new workflow in n8n > Click the three dots menu > Select 'Import from File' > Choose the downloaded JSON file.
2. You'll need accounts with OpenAI, Qdrant, WooCommerce and Google Drive. See the Tools Required section above for links to create accounts with these services.
3. In the n8n credentials manager, create an OpenAI credential: click Create new credential for OpenAI, paste your API key from the OpenAI API page, and save. Select it in all OpenAI nodes.
4. Create a Qdrant credential: for hosted Qdrant, choose HTTP Header or the Qdrant API credential, enter your endpoint URL and API key from your Qdrant dashboard, and save. Map it in both Qdrant Vector Store nodes and the HTTP Request node.
5. Connect WooCommerce: open the WooCommerce REST API settings in your store, create a key with read permission, then in n8n create the WooCommerce credential and enter the store URL, consumer key, and secret. Select it in the personal shopper node.
6. Connect Google Drive: open the Google Drive nodes, choose Create new credential, follow the OAuth screen, and allow file access. Use My Drive or pick the shared drive you need.
7. Configure the Qdrant collection name in both vector store nodes. Replace the placeholder with your own collection. Update the HTTP Request URL so it points to your Qdrant collection delete endpoint.
8. Point Google Drive List to the folder that holds your store docs. Keep file conversion on in the download node so Google Docs convert to text.
9. Verify the token splitter chunk size and overlap match your document length. The defaults 300 and 30 work well for store policies and FAQs.
10. Open the Information Extractor node and review the system prompt and schema. Make sure the fields match your catalog filters such as price, size, brand, and color.
11. In the AI Agent, confirm the tools are connected: language model, memory, RAG tool, WooCommerce tool, and calculator. Save the workflow.
12. Click the manual trigger path to run the HTTP Request once. This clears the Qdrant collection so new embeddings can be written. Then run the ingestion path to load files from Google Drive and push vectors into Qdrant.
13. Test the chat endpoint from the chat trigger with a store hours question. You should get an answer from the RAG system. Then try a product request like bags under 100 to see WooCommerce results.
14. Troubleshoot tips: if no docs are found, check Google Drive permissions and file ids. If vector search fails, confirm the Qdrant URL, API key, and collection name. If product results are empty, verify WooCommerce API keys and that the extracted filters are present.