AI Agent/Assistant Chat Bot RAG

Extension for Store Builder and Multi-Vendor is an intelligent AI assistant that talks to your customers in a chat widget and answers questions about products, content, reviews, documents and order status, using the live data of your store. The same knowledge base also powers an internal staff chat in the back office, with per-group access control. The assistant learns your offer in the background and is available to customers around the clock, so simple questions no longer reach your support team.

Key Features

- A floating chat widget on the storefront that answers customers in their own language, based on your real offer, and can attach the product photo to its answers.

- Automatic synchronisation of products, pages, blog posts, reviews and content blocks into the knowledge base, always in step with the store.

- Your own knowledge documents in PDF, DOCX and TXT, global or pinned to a category, plus public product attachments.

- An internal staff chat over the same knowledge base, with per-document access tags, so sensitive material stays with the right people and never reaches customers.

- Verified order status checking in chat, with a one-time code sent to the customer's email address.

- Knowledge-gap analytics: an unanswered questions panel where one move turns a gap into a FAQ entry or links the question to a product.

- Industry context and a terms glossary, so the same assistant understands the vocabulary of any industry.

This is how the assistant looks to the customer. A chat window opens in the corner of the store, where the customer asks about warranty and invoices, and the assistant replies with specifics and a link to the source document:

Screenshot 1: chat widget on the storefront

The General section is the integration hub. Here you copy the ready REST URLs into your processor, choose the language of the content sent to the AI, and describe your store specialization and industry glossary, which go straight into the assistant prompt:

Screenshot 2: general configuration and industry context

In the Products section you decide which products reach the knowledge base and how they are enriched: with category, brand, market or the full feature set. You also enable loading product documents, hiding the stock quantity, and attaching the product photo to chat answers here:

Screenshot 3: product scope, enrichment and images

The Knowledge documents section sets the document limits, the maximum file size and the number of chunks, and configures the internal staff chat: its n8n webhook and the groups granted full knowledge access:

Screenshot 4: knowledge documents and internal chat setup

The Knowledge documents panel lets you upload your own PDF, DOCX and TXT files. A document without a category is global, with a category it is pinned to that category listing. Each row shows its sync status, its access scope and an Edit action:

Screenshot 5: knowledge documents panel

Every document carries access tags. From the edit form you decide whether it is visible in the customer chat, which admin groups may see it in the internal chat, or whether it is open to all staff, all without re-uploading the file:

Screenshot 6: document access tags

The Order verification section gives full control over what the assistant may show the customer: status, dates, tracking number, item list or value. You enable only what you want to share:

Screenshot 7: order data scope in chat

The Demo mode section protects a public assistant demo with a daily limit, a per-client limit and a security token, so the cost of requests stays under control:

Screenshot 8: demo mode with request throttling

In the Widget section you set the look and behaviour of the chat: its position on screen, its state on entry, the window size on desktop and mobile, and a test link to check the integration before going live:

Screenshot 9: widget configuration

Conversation history collects every customer conversation with the assistant, grouped into sessions, with a content search. It is a window into what customers really ask about:

Screenshot 10: conversation history

The single conversation view replays the full course of a conversation, message by message, so it is easy to judge how the assistant handled the question:

Screenshot 11: single conversation view

Unanswered questions is the knowledge-gap analytics. The assistant recognises when it found no answer, and the panel shows those questions with a counter, from the most frequent to the rarest:

Screenshot 12: unanswered questions

From a gap you write an answer that is saved as a FAQ entry in the knowledge base. After the next synchronisation the assistant starts answering that question:

Screenshot 13: an answer added to the knowledge base

The FAQ list collects answers added by hand. It is a growing body of knowledge that complements the automatically synchronised store content:

Screenshot 14: FAQ entries list

The second way to close a gap is to link the question to an existing product by adding keywords. The product is resynchronised and starts appearing in the assistant's answers:

Screenshot 15: assigning a question to a product

The internal staff chat answers your team from the same knowledge base, but scoped to each employee's role. Sensitive internal documents reach only the allowed admin groups, while customers never see them, and access is resolved live at the moment of the question:

Screenshot 16: internal staff chat

The AI assistant brings together live knowledge of your store, a convenient chat on the storefront, an internal chat for your team and a complete management panel in the back office. It works on CS-Cart and Multi-Vendor, and in the marketplace edition each vendor runs their own assistant. Deployment comes down to connecting a processor to one REST URL, the add-on takes care of the rest.

Feel free to contact us and purchase the add-on!