Build with UbID
Building with UbID means integrating a trust lifecycle, not merely calling an identity endpoint. A successful implementation begins by defining who is authorized to issue evidence, who controls it, who requests it, which policy determines acceptance, and what evidence must remain available after the transaction.
UbID provides a role-based integration model across issuers, holders, wallets, verifiers, relying parties, custodians, and operational services. Open standards provide the exchange contracts; institutional policy provides the meaning, authority, and accountability.
Begin with the trust question
Before selecting a protocol or API, an institution should state the question it needs to answer. Examples include:
- Is this person currently affiliated with an approved institution?
- Does this professional hold a valid qualification for the requested activity?
- Is the presenter entitled to use this credential?
- Is this document-derived claim supported by sufficient evidence?
- Is this device and session suitable for the requested operation?
The question determines the minimum claims, acceptable issuers, assurance level, presentation context, credential status requirements, retention, and final decision owner.
Public integration journey
A typical integration progresses through the following stages:
- Define roles and purpose — identify issuer, holder, wallet, verifier, relying party, controller, processor, and operational owners.
- Design the trust profile — define credential type, claim semantics, accepted issuers, algorithms, status, assurance, disclosure, and policy outcomes.
- Select the exchange pattern — choose approved issuance, presentation, messaging, authentication, or API contracts.
- Discover capabilities — resolve public identifiers and metadata without treating discovery as authorization.
- Prototype with synthetic data — validate the interaction model without using real personal data or production credentials.
- Complete controlled onboarding — agree schemas, environments, security responsibilities, support, and evidence requirements.
- Test positive and negative cases — confirm interoperability, privacy, replay resistance, lifecycle behavior, and failure-closed decisions.
- Approve production use — record the exact profile, versions, owners, controls, and operational readiness.
- Operate and evolve — monitor outcomes, rotate keys, manage credential status, review policy, and coordinate change.
Integration domains
| Domain | Integration objective | Primary UbID products |
|---|---|---|
| Issuance | Convert governed evidence into a signed, deliverable credential | Credential Cloud, Proof, KeyVault, Trust API |
| Holder and wallet | Protect credentials, obtain consent, and produce presentations | Wallet, Access, Recover, Connect |
| Verification | Validate provenance, integrity, status, context, and policy | Proof, Trust API, Connect |
| Trust discovery | Resolve identifiers, public keys, supported profiles, and status | Connect, Trust API, Credential Cloud |
| Authentication and access | Establish a secure user or device session | Access, Wallet |
| Recovery and continuity | Restore controlled access without creating a universal recovery secret | Recover, Wallet, Connect, KeyVault |
| Operations | Observe service posture and retain explainable evidence | Pulse, Sentinel AI |
Stable contracts, replaceable internals
External integrations should depend on versioned contracts rather than on internal databases, container names, implementation libraries, or service topology. The UbID Trust Gateway and product APIs provide a controlled seam between external protocols and internal trust domains.
This separation supports:
- independent evolution of internal components;
- explicit validation at trust boundaries;
- consistent authorization and evidence;
- protocol adaptation without exposing internal state;
- clearer incident and support responsibility;
- reduced coupling between partners and platform operations.
Public documentation and partner documentation
This portal describes the public architecture, participant responsibilities, lifecycle patterns, standards position, metadata concepts, onboarding stages, and publication boundaries.
Controlled partner documentation can additionally provide:
- exact API and protocol profiles;
- environment-specific metadata;
- approved credential definitions and schemas;
- client registration and authorization requirements;
- test vectors and conformance cases;
- support channels and operational service levels;
- change, deprecation, and production approval procedures.
Production credentials, private trust lists, customer configuration, cryptographic material, internal topology, and administrative operations are never distributed through the public portal.
Integration principles
Every integration should preserve the following rules:
- Purpose before data — request only evidence required for a declared transaction.
- Roles before endpoints — assign authority and accountability before selecting interfaces.
- Policy before automation — define what is acceptable before automating a decision.
- Verification before acceptance — cryptographic validity is only one input to institutional policy.
- Consent before presentation — the holder should understand what is requested and for what purpose.
- Lifecycle before launch — define expiry, status, renewal, replacement, revocation, and recovery.
- Evidence without secrets — retain explainable decisions without recording private keys, complete wallet contents, or unnecessary personal data.
- Versioned change — pin profiles and treat unexpected metadata or behavior changes as controlled compatibility events.
Readiness questions
Before requesting partner access, an integration team should be able to answer:
- Which organization is authoritative for each claim?
- Which credentials and claims are actually required?
- What assurance is needed, and how will it be evidenced?
- Who owns the final business or legal decision?
- Which identifiers and metadata must be publicly resolvable?
- How will status, expiry, replacement, and revocation be handled?
- What happens when a dependency is unavailable or a result is indeterminate?
- What data will be retained, for how long, and under which role?
- Which negative tests must pass before production approval?
- Who owns key rotation, incident response, support, and change notification?
Continue with the Integration Model, Issuer Integration, Verifier Integration, or Partner Onboarding.