Wholesale buyers place and change orders in plain language, and see the price they will actually pay.
SynapseAI B2B Wholesale
AI Exchange Hackathon
Dr. Martens On-Site Shopping Assistant
In six weeks, Dr. Martens and Conscia built a prototype on-site shopping assistant for drmartens.com that could search, size, and add to cart.
Designed to make drmartens.com usable by AI shopping assistants: on the site itself, and would be ready for Google's Gemini as that channel opens. Built at the MACH AI Exchange, where MACH Alliance members work on a brand's live business problem.
What the agent does
Designed to understand what a shopper wants, then search, check sizing, and add to cart through the same systems the website already uses.
- Brand
- Dr. Martens
- Built by
- Dr. Martens and Conscia
- Contributed by
- MACH Alliance (Hub operator)
- Industry
- Apparel & Footwear
- Audience
- B2C
01
The problem
Why this exists
Shoppers increasingly start with an AI assistant rather than a website. Being visible to that assistant is not enough. It has to be able to act: check stock, find the right size, add to cart.
Read more
Most storefronts are built only for people clicking through pages. Dr. Martens wanted a build that would make drmartens.com ready for agentic shopping on its own site, and would extend to platforms such as Google's Gemini as they open up.
02
Outcomes
What this demonstrates
- The assistant could search, show product detail, check sizing, add to cart, and review the cart.
- Data and logic are scattered across multiple backend systems, including the CMS, loyalty, the OMS and the commerce platform. For agents to act deterministically, the brand needs to expose a governed capability layer, which is exactly what the orchestration engine does.
- Every assistant action would run through the commerce services the website already uses, so prices, stock, and carts follow one set of rules.
Read more
- The team credits 80% of the groundwork to its earlier move to composable architecture, not to this project.
- A single add-to-cart link would take a shopper from an AI assistant to drmartens.com with the right size in the bag, ready for Gemini.
- Speed is set by the commerce platform, not the AI layer: in one traced add to cart, 5.0 of 6.0 seconds were the commerce engine.
What you would add to ship it
- Google access for Gemini as a channel. It is waiting on Google.
- Markets beyond Great Britain. The proof of concept covers one.
- A conversion and order value baseline, so the assistant's effect can be measured.
03
How it works
Architecture
A shopping assistant embedded in the storefront would use two kinds of tool. Commerce tools (sizing guide, search, product detail, add to cart, view cart) would be MCP tools on the Conscia platform, each backed by one flow in Conscia's orchestration layer, which in turn calls the commerce platform. Page tools (navigate, page context, category tree) would act on the page the shopper is viewing.
Read more
The same orchestration layer already serves the web store, the mobile app, and in-store kiosks, so AI assistants would be added as a fourth channel rather than a new build.
For external assistants such as Gemini, products would surface from the existing merchant feed, and one add-to-cart link would hand the shopper back to Dr. Martens' own checkout.
Watch the demo (2 minutes, recorded by Conscia).
Open standards used
Model Context Protocol (MCP)REST APIs
Tech stack
ConsciaSAP Commerce Cloud (OCC)Next.jsUniversal Commerce Protocol (UCP)Google AI Shopping
Everyone credited on this solution
Next step
Request an intro
This one is the MACH Alliance’s own work. Ask about making your storefront buyable by agents, and the people who built it will answer.
Not ready to talk to anyone? MACH Insider is the Alliance mailing list. New solutions, program openings, and the research behind them, once a month.
Related solutions
Every promoted item is checked against the source of truth, with problems surfacing in minutes, not weeks.
Wegmans Marketplace Validation Agents
AI Exchange HackathonMigration mapping that ran 6 to 9 months now finishes in hours, with the write-up produced alongside.
DeepShift Migration AI
AI Exchange Hackathon
Verification
Contributor submission
The detail on this page comes from the contributor's own account of their own work. The MACH Alliance reviews a submission before publishing it and does not independently verify the figures in it.
What this page does not establish
- No conversion or average order value baseline was published for this build.
Where this came from
Written from the Dr. Martens and Conscia session deck at MACH X Amsterdam, September 29–30, 2026, its architecture slides, and the recorded demo. Scoped to Great Britain. A separate Dr. Martens hackathon entry covers migration mapping.
Status: Reference prototypeAdded Oct 15, 2026A listing is an editorial decision, not a MACH Alliance certification.
Built something worth showing?
Put your agentic solution in front of the enterprise leaders who use the hub to decide what to build next. Submissions are open to MACH Alliance Members, Collaborators, or Ambassadors.