Product · The Financial Action Control Plane
ValarIQ is the Financial Action Control Plane.
ValarIQ combines policy, behavioural intelligence and verifiable evidence to authorise financial actions before they execute, whether initiated by people, workflows, APIs or AI agents.
ValarIQ authorises; customer systems execute.
The customer problem
Finance teams need more straight-through automation without losing control. Human review on every transaction does not scale. Unchecked automation creates out-of-policy risk. ValarIQ sits between proposed financial actions and execution so organisations can automate safely as people, workflows, APIs and AI agents initiate more of the work.
- Increase straight-through decisions inside policy
- Reduce unnecessary human reviews
- Control exceptions with named approval
- Stop prohibited actions before they reach rails
- Keep attributable decision history
- Progress safely toward agentic finance
Control and Intelligence
Control
Identity, authority, policy, decide, prove
Define who may act, what they may do, and return ALLOW, ESCALATE or DENY before customer systems execute.
Intelligence
Profiles, verification, confidence
Behavioural profiles and evidence confidence make verification proactive: clear what is normal, hold what is not.
How Control and Intelligence work together
Product view of the decision boundary. For the full initiator-to-execution diagram, see Architecture.
Control
Who may act, and within what limits
- Identity
- Authority
- Policy
- Decision
- Evidence
ALLOW · ESCALATE · DENY
Customer execution
Your rails stay in control of money movement
- Payment provider
- Bank
- ERP
- Workflow
Intelligence
Behavioural profiles and evidence confidence sit alongside Control so routine actions clear faster, and exceptions surface with context before execution.
Path: Identity · Authority · Intelligence · Policy · Decision · Evidence
ValarIQ authorises. Customer systems execute. Full architecture →
Core capabilities
Buyer view of principal capabilities. Detailed technical register available for diligence.
Identify and govern
Actor identity registry
OperationalKnow which people, systems and agents can propose financial actions.
Not a full enterprise CMDB. Discovery promote is separate from authorisation.
Authority profiles
OperationalDefine what each actor may do: amounts, destinations, actions and owners.
Financial authority deepest; non-financial templates limited.
Authorise and decide
Action Gateway
OperationalReturn ALLOW, ESCALATE or DENY before actions reach execution systems.
ValarIQ does not execute payments or hold funds.
Policy evaluation
OperationalApply transaction policy bands and controls at decision time.
Core refund defaults are operational via the Action Gateway. Expanded governance catalogues deepen with customer programmes.
Intelligence and verification
Context Assurance
OperationalEvaluate evidence confidence so decisions are proactive, not retrospective.
Deterministic core is operational. External evidence providers require integration. Does not independently guarantee that source facts are true.
Behavioural intelligence
OperationalBuild profiles from observed actions so verification focuses on what looks off.
Templates require tenant adoption. Playbooks do not self-certify readiness. Mappings do not imply compliance. Recommendations are SUGGESTED — not predicted ROI. Patterns are tenant-specific observations — no benchmarks. Provider reliability and execution outcomes remain Configurable without activated providers.
Evidence and continuing assurance
Signed evidence
OperationalRecord attributable decisions with integrity checks within implemented scope.
Signed decision evidence is operational within implemented scope. Broader relationship traversal is provided by Evidence Graph.
Evidence lifecycle
OperationalKeep historical truth fixed when current authority or applicability changes.
Operational for signed decision evidence. Completeness depends on what was captured when the decision was sealed. Confirming that a payment rail actually executed remains a separate capability. Re-authorisation after change requires a complete agent and policy setup.
Integrate and operate
Control API and gateway client
ConfigurableIntegrate decide and verify into customer applications, workflows and agent runtimes.
Gateway client is available for integration programmes. Not distributed as a public package-registry release.
Workspace tenancy
OperationalIsolate customer environments with roles, membership and onboarding controls.
Workspace lifecycle and onboarding are operational with logical isolation. Dedicated residency is roadmap.
Product layers
Agent Registry
What agents do we have? Who owns them?
Authority / Mandate
What can it do?
Action Gateway
May this action proceed?
Policy Engine
What controls apply?
Context Assurance
Is the supporting context trustworthy?
Evidence & lifecycle
Can we prove what happened, and whether it still holds?
Monitoring & response
What is going wrong?
Executive intelligence
What requires a decision?
Observe-to-Enforce deployment
1
Define one financial action
Choose a workflow with clear policy and measurable exceptions.
2
Connect in Observe mode
Evaluate decisions without blocking production execution.
3
Configure authority and policy
Set identity, limits, destinations and escalation paths.
4
Review decisions and exceptions
Validate outcomes with Risk and operations.
5
Approve Enforce mode
Block prohibited actions and require approval where needed.
6
Expand across workflows
Reuse the same control plane as automation and agents grow.
Enterprise boundary
- ValarIQ authorises; customer systems execute payments, ERP and workflow actions.
- An ALLOW is authority to proceed, not confirmation that execution occurred.
- Evidence proves the integrity of the recorded decision within implemented scope.
- Missing, stale, contradicted and unknown evidence remains visible, never silently filled with favourable assumptions.