CirculeID

Infrastructure

Four moving parts, and one that has to outlive the contract

A resolver, a record store, an event store and a credential layer. The architectural question that matters is not throughput — it is whether an identifier issued today still resolves when the product reaches end of life in 2041.

Resolver
GS1 Digital Link
Events
Append-only
Residency
Selectable, EU available

Definition

What does Digital Product Passport infrastructure consist of?

A passport platform has four components: a resolver mapping GS1 Digital Link identifiers to role-scoped responses, a record store holding passport data, an append-only event store holding EPCIS 2.0 supply chain events, and a credential layer issuing and verifying W3C Verifiable Credentials. Integrations sit around those four.

The separation is not incidental. Reads follow public scans and writes follow machine-generated events, and the GS1 Digital Link resolver has to answer the first without waiting on the second.

Architecture

What each layer is responsible for

Knowing which layer owns which guarantee is what makes an architecture review with your platform team short.
Passport platform layers, their responsibility and the standard each follows
LayerResponsibilityStandard
ResolverTurns an identifier into the response the caller is entitled toGS1 Digital Link
Record storeHolds the current passport state and its access policyCIRPASS-aligned model
Event storeAppend-only history of what happened to each objectGS1 EPCIS 2.0
Credential layerIssues, verifies and revokes signed claimsW3C VC 2.0 · DID
IntegrationMoves data in from ERP, PLM and partner systemsREST · webhooks

Answers

Frequently asked questions

What are the moving parts of a passport platform?

Four. A resolver that turns a GS1 Digital Link identifier into the caller-appropriate response, a record store holding the passport data, an append-only event store holding EPCIS 2.0 events, and a credential layer that issues and verifies W3C Verifiable Credentials. Everything else is integration around those four.

Does this run on a blockchain?

Not for the passport record itself, which would be both slow and inappropriate for data that must sometimes be corrected or erased. A distributed ledger is used only where anchoring adds something a signature does not — notably tamper-evident timestamping and revocation registries that no single party should control.

Where is our data held?

Data residency is selectable per tenant, and EU-resident deployments keep passport records, events and credentials within the EU. This matters more than usual here, because passport data is regulated product data that a market surveillance authority may need to reach for the lifetime of the product.

What has to keep working for fifteen years?

The identifier must keep resolving, and the credentials must keep verifying. Both are addressed by design rather than by promise: identifiers follow GS1 Digital Link and can resolve on your own domain, and credentials verify against Decentralized Identifiers that do not depend on this platform continuing to exist.

How does load actually behave?

Read volume follows scans, which are bursty and public-facing, while write volume follows supply chain events, which are steady and machine-driven. Those are different workloads, so the resolver and the event store scale independently rather than sharing one bottleneck.

Next step

Bring your platform team

We will walk the architecture, the residency model and the exit path, and answer the question that actually decides it: what happens if you stop using us.

Index