Test ideas in real workflows
Try AI advice, new shopping experiences and automated processes. The measure: does the idea actually help merchants and customers?
Vendune is Stefan’s experiment platform for the next generation of e-commerce. AI, new shopping experiences and merchant workflows can be explored directly in a working commerce system.
Working prototype · developed in public

Vendune is neither a competitor nor a replacement for Shopware. It is a development workshop where new products, services and approaches to commerce can take shape. Ideas that prove useful can inspire Shopware.
Try AI advice, new shopping experiences and automated processes. The measure: does the idea actually help merchants and customers?
Modern technology and public development can attract additional developer communities. Contributions, extensions and actual use matter more than stars alone.
Offer Shopware Payments through new products, standalone experiences and agentic workflows. Its provider integration remains a separate private service.
Storyfront shapes the shopping experience. Vendune keeps commerce data authoritative. Each studio owns its part, connected through the same integrated interfaces.
Create a shop and manage it in Vendune Studio.
experience.vendune.aiConnect brand and catalog, edit in the original Storyfront Studio and publish.
Products, variants, prices, stock, customers, taxes, cart and orders.
Original Studio & rendererBrand, knowledge, questions, scenes, design, preview and publication.
Presentation and commerce stay connected: Storyfront can shape product stories, but editable narrative text cannot overwrite prices, stock or orders.
“Prepare my new collection for France.”
Products, documents, shop knowledge and existing workflows provide context.
AI can prepare content, translations or actions. Knowledge and proposals do not mean an autonomously trained, all-purpose intelligence.
Changes pass through permissions, preview and approval. Prices and orders remain authoritative in the core.
Two real screen tours. English, with burned-in captions and no sound. Chapters jump straight to the relevant moment.
Onboarding, original Storyfront Studio, connection to Vendune and order completion in the embedded checkout.
The video could not load. Open the download link or reload this page.
The film edits a new draft using an existing AI demo catalog, then explicitly switches to Cobalt Atelier for shopping. It shows a simulated order, not a real charge or new image generation.
Create a shop with sample products, then explore administration, orders, shop knowledge, apps and the storefront.
The video could not load. Open the download link or reload this page.
Atelier North was newly created in the film with twelve demo products. The admin tour then uses the already authenticated Cobalt Atelier shop. The recording does not verify receipt of an access email or setting a new password.
Choose the route that fits your question. Use your own test data and check that payments are configured as simulations.
Create an account or sign in. Choose an empty shop or sample products. Edit a product in Studio and open the shop.
Try a shopProvide a brand and reviewed catalog: images, CSV/Excel, product URLs or proposed AI demo products. Review the import, edit in the original Studio, test the preview and deliberately publish.
Try an experienceAdd a demo product, open the cart and finish the simulated checkout. Check the order and customer association in that shop’s Vendune Studio.
Check a purchaseExisting public experiments may use different generations of the renderer. The tour provides a traceable reference for the current native connection.
From shops to platform operations. Protected areas require an authorized personal account.
The classic SaaS entry without automatically provisioning Storyfront.
start.vendune.aiOnboarding and access to the connected original Storyfront Studio.
experience.vendune.aiManage an existing shop; choose the relevant shop after signing in.
app.vendune.aiOperator area for shops and platform operations. Authorized team members only.
app.vendune.aiCommerce core, changes, issues and contributions on GitHub.
github.comUnderstand architecture, features, interfaces and boundaries.
sthamann.github.ioPeople and coding agents work on the same versioned app contract.
github.comMathematically checked properties of selected critical rules.
github.comAI connects to products, documents, shop knowledge, apps and workflows. A merchant describes a goal and the system prepares concrete proposals. It does not mean unlimited self-learning intelligence or uncontrolled changes.
Storefronts, apps and agents use the same commerce functions. Prices, stock, taxes, permissions and orders are checked centrally. Rust alone does not prove a particular speed improvement; that requires comparable measurements.
App Studio can describe data, API, admin, storefront and AI features. Versioned packages can be tested in a private environment and rolled out selectively. External apps remain language-independent; the core stays authoritative.
No. Vendune is a working prototype. Broad production readiness, community impact and each external provider path must prove themselves separately. Verified demo workflows do not establish real payment settlement.
The public core provides a general payment-provider contract. The specific Shopware Payments implementation remains private and needs authorized provider configuration, merchant assignment and verified onboarding. A demo order does not confirm a payment.
Lean checks properties of selected critical rules, such as stock or refund limits. These bounded proofs do not replace integration tests or guarantee a completely bug-free system.
Bring a merchant task, a shopping idea or an app idea. Test a concrete workflow and tell us where it helps — and where it still gets in the way.
Start with a shopExplore the project