What This Workflow Does
This workflow creates a chatbot named “Pizzaro” that helps customers order pizzas by chat.
It solves the problem of manual and slow chat handling in busy pizza shops.
The chatbot answers menu questions, takes orders, and gives order status updates automatically.
Here, the chatbot uses AI to understand messages and talk naturally with customers, saving time and reducing errors.
Who Should Use This Workflow
This automation is for pizza shop managers or teams who get many chat orders and questions daily.
It helps staff stop juggling chats manually and avoid order mistakes.
If customers ask about menus, place orders, or check status through chat, this workflow can handle all that without delays.
Tools and Services Used
- Chat Trigger node: To receive customer chat messages in real time.
- Window Buffer Memory node: To keep conversation history for smooth replies.
- HTTP Request nodes: Three nodes to get pizza menus, submit orders, and check order status using backend APIs.
- AI Agent node: Acts as the brain to decide user intent and choose the right action.
- Calculator node: Calculates totals like prices and quantities.
- Chat OpenAI node: Produces human-like chat answers using OpenAI’s language model.
Inputs, Processing Steps, and Outputs
Inputs
- Incoming customer chat messages with questions or orders.
- Access to pizza menu and order status APIs.
- OpenAI API key for generating replies.
Processing Steps
- The Chat Trigger node receives chat messages from customers.
- The Window Buffer Memory stores recent conversation to keep context.
- The AI Agent analyzes the message to understand if the user wants the menu, to order, or check status.
- Depending on intent, the AI Agent calls:
- Get Products (HTTP GET) node to fetch the current pizza menu.
- Order Product (HTTP POST) node to send the order details.
- Get Order (HTTP GET) node to get order status.
- The Calculator node is used to compute total prices or quantities if needed.
- The Chat OpenAI node generates clear and polite replies to send back to the customer.
Outputs
- Quick, accurate chat replies about menu items.
- Order confirmations with details and thanks.
- Live updates on order status showing pizza type and quantity.
- Reduced manual work and errors for the pizza team.
Beginner Step-by-Step: How to Use This Workflow in n8n Production
Step 1: Import the Workflow
- Download the workflow file using the Download button on this page.
- Open the n8n editor where you want to run the chatbot.
- Choose “Import from File” and select the downloaded workflow file.
Step 2: Configure Credentials and Settings
- Set your OpenAI API key in the Chat OpenAI node credentials section.
- Update URLs or IDs if needed in the HTTP Request nodes (Get Products, Order Product, and Get Order) to match your pizza shop’s backend API.
- Check the Chat Trigger node has the webhook URL public so it receives messages from customers.
- Review the initial greeting text if you want to add your own welcome message.
Step 3: Test the Workflow
- Send a sample chat message to the webhook URL such as “What pizzas do you have?” and see if it replies with the menu.
- Try placing a test order message like “I want 2 Margherita pizzas, name is Sarah.” and check confirmation.
- Ask for “Order status for Sarah” and verify order updates are sent back.
Step 4: Activate the Workflow
- After successful testing, activate the workflow by toggling it ON in the n8n editor.
- Confirm the chatbot is live and responds to customer chats continuously.
If using self hosting n8n, ensure the server stays online and reachable to keep customer chats flowing.
Customization Ideas
- Change the welcome message in the Chat Trigger node to fit your brand voice.
- Update the backend API that provides pizza menu data to add or remove items without changing the workflow.
- Edit the AI Agent’s prompt to change how order confirmations sound.
- Add tax, discounts, or delivery fees calculations using the Calculator node.
- Extend the chatbot to answer questions in multiple languages by changing the Chat OpenAI node prompt.

