One governed assistant for your storefront and your back office.
A single adapter serves both shoppers and merchants. Persona is just the user’s role and permissions — so a customer can track an order while only staff can issue a refund or adjust stock.
Control is the default here.
One adapter serves both storefront customers and back-office merchants; the persona is nothing more than user.role plus permission and available-action references. A customer tracks an order; the same code path, with a merchant’s references, exposes approval-gated refunds and stock adjustments.
Whether you run Shopify (hosted, OAuth2, Admin API token held in the adapter) or nopCommerce (self-hosted, JWT, C# services), the difference lives inside the adapter and is invisible to the platform core. Swapping one for the other doesn’t touch anything above the adapter.
This is a documented integration approach with worked examples — not a pre-built, install-and-go connector. Ask us what a production integration looks like for your setup.
Every action an agent takes here runs through the same gate: least-privilege permissions, policy and an audit trail, with the sensitive steps held for human approval. The controls behind it are set out on the security and data-handling page.
What it looks like in practice.
Illustrative examples of what a host would configure — tool ids and risk levels are the host’s to define. The primitives underneath are built and tested.
Storefront order-status & returns
With user.role = customer, an agent tracks the order read-only and initiates a return through approval — the refund tool simply isn’t in scope.
Merchant refund & stock adjust
Flip the role to merchant and issue_refund and adjust_stock appear as approval-gated actions on the very same adapter — elevated, but still approved and audited.
Catalog & promotions report
A merchandiser asks which SKUs are slow-moving and which promo drove the most revenue; specialist agents return a table and a promo-revenue chart, store-scoped.
Financial services & fintech
Every action is the intersection of what the user, the agent, and org policy permit — fail-closed, approval-gated for anything risky, and recorded to an append-only audit.
Industry · LogisticsLogistics & supply chain
Agents surface what’s at risk, weigh the options, and route reroutes, reschedules, and discrepancy notes through approval — with a full audit trail.
Industry · HR & workforceHR & workforce
From self-serve policy answers to people analytics, agents read only what the user is allowed to see, keep humans in the loop on decisions, and expose aggregates without leaking individual rows.
See governed AI for E-commerce.
Book a demo and we’ll tailor the walkthrough to your sector — the governance model, the two-seam integration, and an example close to your systems.