Commerce In development
Project in development
Commerce OS
Concept UI · Development preview · No customer or production data
Products that describe themselves. A store that answers its shoppers.
- Built for
- Online brands and commerce teams
- Industry
- Commerce
- Project Type
- AI-native commerce platform
Description
Commerce OS is an AI-native platform for running online stores, in development. It covers the store itself, the catalogue, orders, customers and payments, with AI doing the repetitive work around them.
Upload product photos and the AI writes the product copy from the images and details, ready for a person to review and publish. On the storefront, an AI shopping assistant answers shopper questions about products and orders. Every product page is built to be found, with search and AI-answer metadata generated automatically.
People stay in charge of what goes live: AI drafts, the team approves.
Deployment options
Available as a subscription or as a deployment on your infrastructure, subject to the agreed scope. Forward Deploy retains the software’s intellectual property; usage rights, fees, support and running costs are agreed in writing. Read the terms
Specifications
- Status
- In development
- Focus
- Storefronts, catalogue, orders and AI assistance
- AI
- Product copy from photos; storefront shopping assistant
- Built for
- Online brands and commerce teams
- Application
- Next.js, TypeScript and PostgreSQL
- Release
- Not publicly released
Features
AI product copy
Descriptions, features and tags written from the product’s own photos and details.
AI shopping assistant
Answers shopper questions on the storefront, around the clock.
Search-ready pages
Product pages ship with search and AI-answer metadata, sitemaps and structured data.
Orders & fulfilment
Orders, partial fulfilment and refunds tracked with clear status rules.
Human approval
Nothing AI-written goes live until the catalogue owner approves it.
Architecture
Commerce OS is being built as one platform for storefronts and store operations. AI assists with product content and shopper questions, while approval gates anything that can be published. This diagram shows the intended operating flow, not a deployed system.
Explore the architecture ↗ (new tab)Expected Impact
Expected outcomes, not measured results. Figures will be added once the product is piloted and compared with a baseline.
Faster listing
Products live sooner
AI-written product copy is expected to cut the time from photo to published product.
How we’ll measure it Track preparation time per approved product.
Better answers
Shoppers helped instantly
An AI assistant on the storefront is expected to answer routine questions without staff.
How we’ll measure it Track questions answered without a support ticket.
More traffic
Pages that get found
Automatic search and AI-answer metadata is expected to improve discovery of product pages.
How we’ll measure it Track organic visits and AI-answer citations per product.
More control
Clear approval before publishing
A shared approval step is expected to keep AI-written content accurate.
How we’ll measure it Track rework and review cycles per product.