A chair, a lamp, a table in one chat: a

TL;DR: Customers shop in baskets, not individual items. aiSTAFF tracks a list of multiple items in a single chat, taking each item to view, review, refine, select, skip, or exchange, and ending with an order summary and running total. There is no card payment in the chat: the order is product discovery plus a call.

One Item Bot Problem

Most chatbots in a store can find a single product. Ask for a seat, get a seat. But the client who is designing a room doesn't want a chair, they want a chair, a lamp and a table, in one conversation, with a complete feeling before they put it together. A bot that processes one item at a time forces the shopper to start over with each item, loses track of them, and never provides a cart view. The conversation feels like three separate quests instead of one shopping trip. If you're selling online, our AI Sales Bot handles the entire cart and here's how the cart logic works.

The cart is a small state machine

aiSTAFF keeps a live shopping list for conversation and keeps track of where each item sits. All elements move through several states and the bot knows which one is active.

Browsing. The customer states a need (“I want a reading lamp”) and the bot returns the corresponding cards.
Consideration. The buyer asks for one option: price, color, whether it will be darkened or not.
Refinement. "cheaper", "warm light", "something higher". The bot will search again in the same item slot.
Choose. The customer selects an option and it is added to the cart with its price.
Skip. They leave the element for now; The bot moves without forgetting the rest.
Exchange. They will replace the selected item with another item and refresh the whole.

Since the bot keeps the entire list in memory, it can at any point respond to "what I have so far" and recalculate the running total. Conversation memory levels, set by the plan, decide how far the bot can see, which is important in long chats. The customer experiences one continuous journey, not a series of disconnected searches.

Running Total, Honest Saved

Each selection adds its actual catalog price to the running total, and each exchange or omission adjusts it. The numbers come directly from the catalog, so the totals reflect current prices and discounts, not guesswork. When the customer is finished, the bot summarizes: each item, its price and the total amount in the cart, how the cashier reads your order. The items themselves come in the form of full cards with price, old price, rating, and inventory covered in product cards in chat that converts. Availability is checked before anything is confirmed, so an out-of-stock item is never included in the cart, which is detailed in Availability Checks.

Honest limit: no payment in chat

That line does not cross aiSTAFF here. Cart is for discovery, not checkout. There is no card entry, no charge, no payment button inside the conversation. "Order" means that the bot has collected the cart and now transfers the customer to the actual next step: call, contact or saved call with a phone number and a list of items. A teammate or follow-up message closes the sale.

It's intentional and it's the true way to sell in chat. A prospective order that a bot can't deliver creates a wasted moment at the worst possible time, right when the customer is ready to buy. Instead, the bot captures intent and cart, then offers a call to arrange pickup, delivery, or payment without a chat. Lead capture can be passive (only when the user offers a number) or proactive (offered when the intent to buy is high), and once captured, the bot continues to help, not silently. An overview of the selling engine is in the hub, AI Chatbot That Sells Your Catalog.

grows the basket as a good assistant

The multi-point shopping cart is a natural place to increase the value of the order. When a user selects a hammer, the bot can suggest nails and gloves from your own catalog, rather than random products, because it understands related items. This is how a single item visit turns into a three-point cart explained Bot Offers Brackets. The offers stay in real stock and the relevance gate keeps them honest, so the bot never fills the cart with an item you don't carry that is covered by the relevance gate. Contextual ranking means "Show me cheap" sorting by price, and gift requests are based on ratings, detailed Show me cheap.

Working example

A customer opens a chat with a furniture store. "I'm setting up a home office." The bot asks what they need and they list a table, a chair and a lamp. He goes to the desk first, returns three, the client improves to "600 GEL", chooses one. Bench: Added two options, change from first choice to mesh model. Lamp: The user will skip it for now. The bot offers a cable tray that pairs with a desk; they add Then sums up: table, chair, cable tray, with a running total and asks for a number to arrange delivery. The customer gives it. No payment was made in the chat, but the link for the three units with the phone number was in the inbox, ready to be closed. The ROI side of this is ecommerce chatbot ROI for Georgian store and the Georgian customer could conduct this entire chat in their own language, which is covered by How Georgian customer buys your English catalog.

FAQ

Can a chatbot manage more than one product at a time?

Yes. It keeps a live shopping list and goes through each product, previews, reviews, refines, selects, skips or changes, then sums up the entire cart with a running total at the end.

Can the user pay within the chat?

No. Cart is for product discovery. The bot builds a cart and delivers a callback, contact, or saved lead to close the sale from the chat. Payment by card is not possible while talking.

How does a running total stay accurate?

Each selection adds to the actual catalog price and changes or omits the total quantity adjustment. Prices are from a synchronized catalog, so the figure reflects current prices and discounts, not an estimate.

Do you offer additional items?

Yes. The bot can suggest related products from your own catalog, such as hammer nails to raise the basket. Offers remain in actual stock and pass the compliance gate.