Previous PayReality technical work. This page documents an earlier product direction and does not describe our current offering or availability. See what we're working on now.
One authority-to-outcome workflow.
PayReality brings approved business rules, delegated authority, and trusted current facts to a decision about one exact action. Where the customer's execution system is integrated, it applies the action-bound authorization and returns a report that can be compared with what was approved. Each named capability below supports that single journey; the customer does not buy five unrelated products.
Every product answers a different part of one question
"For this exact action, what authority applies now, what may execute, and what did the destination report?" isn't answered by a single check. It requires knowing who holds delegated authority (the Authority Graph), what conditions apply to a given action (Runtime Policies), a deterministic way to combine both into one outcome (the Runtime Authority decision engine), and a record of what was decided and why (the Evidence Portal and Authorization Receipts). This page is the catalog: what each capability does, and where to explore it on its own. For how they evaluate an intent together, see Platform.
End to end, this is a lifecycle, not a one-time setup: authority is proposed from your organizational sources and only takes effect once a human has approved the exact version; current cross-system facts are resolved; the exact action is evaluated against your approved authority and policy; human approval is obtained where your governance requires it; a constrained authorization is issued; a compatible enforcement point applies it; and, where a trusted destination reports back what it did, the reported execution is reconciled against what was actually authorized. The authority lifecycle stays auditable throughout, not just at the moment of decision.
Not five tools. One chain.
Explore each capability
Runtime Authority
The decision engine at the center of the runtime: evaluates each AI agent's signed intent against your organization's delegated authority and returns an authoritative decision before execution.
Authority Graph
Models who holds delegated authority, and who it's been extended to, from your existing governance documents.
Runtime Policies
Compiles delegated authority into deterministic, versioned rules a runtime can evaluate.
Evidence Portal
The enterprise evidence layer: every decision, signed, searchable, audited, and exportable, including reconciled execution outcomes where a trusted destination reports back.
Authorization Receipts
A signed receipt for every decision, retrievable today. A portable, independently verifiable artifact is the roadmap beyond that.
One intent, evaluated end to end
AI agent attempts an action
Trusted observation (agent-signed, or an Adapter-reported operation)
Authority Graph and Runtime Policy evaluated
Decision: Allow / Deny / Human Review
Human approval, where the decision requires it
Capability Authorization, where the action requires it
Compatible enforcement point applies the authorization
Reported execution reconciled against what was authorized, where a trusted destination reports back
Evidence and Authorization Receipt preserved
Human approval and Capability Authorization aren't part of every decision: they apply only where a decision escalates to Human Review, or where a use case is configured to require a Capability before a compatible enforcement point will apply it. An Allow decision that needs neither resolves straight to Evidence. Reconciliation is conditional too: it only runs once a separately authenticated destination or trusted execution adapter reports execution back, and even then it confirms what that trusted party reported, not the underlying business or real-world truth independently.
For the full request and response detail behind this flow, see Runtime API in the Developers section, or Architecture for how each component fits together.
Affected pending actions and ambiguous post-dispatch outcomes are areas we are testing. Within connected workflows, the intended control identifies actions needing revalidation and keeps a missing outcome open for investigation without silently granting a second execution. This is planned functionality; its reach depends on integrations and destination evidence.
Start with the flagship product
Runtime Authority is where every submitted intent is actually evaluated: the best place to start.