AI-driven operation
Load and fix the catalog by talking
Ecommerce you run by talking to it
Catalog, pricing, inventory and orders on the same platform as your site, with an AI layer that loads, corrects and audits them. Your team asks in plain language and the platform acts on the real catalog.
Running the catalog is a conversation
The platform's assistant works on the real catalog: it creates products, fixes prices, completes records and answers questions about what is loaded. It does not draft text for someone to copy, it executes.
Bulk product creation
Hand it the list your team already has, in whatever format it is in, and it loads the products.
Dictated corrections
Fix a price, adjust inventory, change an order status or complete a product's SEO fields, in one sentence.
Questions about the catalog
Which products have no image, which were left uncategorized, which are below the minimum inventory level.
It never invents the form
It reads the real product record definition before writing: fields, types, required flags and valid values. If a value does not fit, it catches that before saving.
Under the CMS rules
It works inside the platform's permissions and validations, and every change is logged like any other edit.
MCP server: your catalog inside the AI you already use
The platform exposes its capabilities over MCP, the open standard that connects AI assistants to outside systems. Your team can work on the catalog from the assistant it already uses, while building a campaign or reviewing a report, with no exports and no custom integration to rebuild every time a tool changes.
Open standard
Any MCP compatible assistant connects and works on the catalog. This is not a closed list of vendors: it is a standard.
Migrate an entire catalog
Bring products, brands and categories over from another system, reviewing and correcting in the same pass.
Audit what is already there
What is missing, what is out of date and what is duplicated, across the full catalog and not a sample.
It grows with the platform
Every new module adds its capabilities to the assistant automatically. There is nothing to reconfigure separately.
Multi-brand: one catalog per brand, one platform
Every brand in your organization runs its own catalog, its own pricing and its own orders, with no duplicated installs and no extra licenses. The platform resolves on its own which brand each visit and each content request belongs to.
A catalog per brand
Products, brands and categories of its own for each one, with no team seeing or touching another team's catalog.
Automatic resolution
By the incoming domain or by the content section: nobody picks the brand by hand.
Orders keep their source
Every order and every inquiry is recorded with the brand it came in through.
One single update
One deployment, one place to look when something breaks, and as many brands as you need.
Headless: the catalog feeds any channel, not just the site
This is not an off the shelf storefront. The catalog lives on the same platform as your site content and is exposed over API, so a product loaded once serves the site, an app or a campaign landing page without loading it again.
One source of truth
A product is loaded once and published to every channel. There are no two catalogs to sync or to audit.
An API for any front end
The site, a native app or a campaign landing page consume the same catalog, each with its own interface.
One change, every channel
A price or inventory fix lands everywhere at once, with no sync jobs to run.
Catalog and content together
The product record lives alongside the site pages: they link to each other and are edited in one place, not in two systems.
Catalog SEO, product by product
Every product gets its own optimization record, integrated with SEO Pack. And the bulk review, which is exactly the work nobody does by hand, is handled by the assistant.
SEO fields per product
Title, description, social image and product structured data, in one tab of the same form.
Sitemap with collections
Products enter the sitemap on their own, one entry per item, and it regenerates when you publish.
Visible to AI search
The llms.txt file exposes the catalog to the assistants that now answer instead of Google, rather than leaving them to guess.
Assisted bulk review
Filling in the metadata for a whole category becomes one request, not an afternoon of work.
The product record: all the detail, in tabs
Under the AI layer there is a record with the detail today's commerce demands, organized in tabs so filling it in stays fast even with a lot of information. The assistant completes it faster, and your team edits any field by hand whenever it wants.
Data and selling units
Name per language, SKU, friendly URL, brand, category, base price and stock status, with conversions to sell by the box or by the square foot.
Descriptions and media
Short and long description per language, main image, gallery, video and blocks for spec sheets, drawings or room shots.
Logistics and identifiers
Weight, dimensions, shipping cost and lead time, low stock threshold, plus GTIN for marketplaces and links to certificates or manuals.
Offers and recommendations
Promotional price with start and end dates, a label per language, and upsell and cross sell products.
Gaps, and the spreadsheet if needed
The assistant finds incomplete records before they go live. And if your team prefers CSV, import and export are still there.
Brands and categories: how the catalog is organized
A well classified catalog is easier to browse, easier to filter and easier to rank. And it gives the assistant the vocabulary it needs to understand what you are asking about.
Brands with their logo
Name per language and logo. Each product is tied to a brand, and that enables filters and brand pages.
Category tree
Categories, nested subcategories and an image per category: the structure a visitor walks to find what they came for.
The basis for every listing
The site's grids and filters are built from this classification, with no separate setup.
Cart and orders: from interest to purchase
Products are added to a cart and, once it is confirmed, it becomes an order. Your team manages it from the panel, and can also look it up and update it by asking the assistant.
Cart
Built from the site by adding products and quantities, with subtotals calculated for you.
Order confirmation
Checkout moves the cart into confirmed order status.
The order record
Reference, status, buyer details, total and dates, with line by line detail for every item.
Managed from the panel
Your team sees every order, looks it up and updates its status without leaving the platform.