Runtime policies
29
Enterprise Judgment Runtime - IRIS
Policy intelligence, controls, procedures, system configuration, historical decisions, and SME knowledge
Judgment runtimePolicy intelligencePromotion gates
Enterprise Judgment Runtime
Most enterprise judgment is not stored in one policy repository. It is distributed across policy documents, controls, procedures, SAP configuration, workflow rules, historical decisions, audit findings, exception practices, and SME knowledge. The runtime makes that judgment visible, structured, governable, and executable.
Policy assets
457
Candidates
428
Controls
13
Historical decisions
5,422,406
Readiness
47%
Policy Intelligence Layer
The runtime should not replace SharePoint, Confluence, SAP, Ariba, GRC, or control repositories. It should create a governed policy intelligence layer above them.
| Source layer | SharePoint, Confluence, SAP, Ariba, GRC, controls, AI governance |
|---|---|
| Registry | Policy metadata, owner, source, version, effective dates, approval status |
| Knowledge graph | Policies linked to controls, processes, roles, systems, agents, and evidence |
| Runtime | Policy evaluation during agent and human execution |
| Replay | Decision proves which policy version and evidence were used |
Enterprise Judgment Layer
Policy documents are only one source. The runtime also needs embedded, learned, and human knowledge to represent how finance actually operates.
| Explicit knowledge | Policies, controls, procedures |
|---|---|
| Embedded knowledge | SAP rules, workflow configuration, approval hierarchies |
| Learned knowledge | Historical decisions, exceptions, outcomes, overrides |
| Human knowledge | SME guidance, controller expertise, tribal knowledge |
| Promotion gates | Owner, version, control mapping, evidence, replay, and human accountability |
Live Enterprise Judgment Runtime Audit
Live runtime
Runtime policies
29
Policy assets
457
Candidates
428
Judgment readiness
47%
Controls
13
Procedures
18
System configs
0
Historical decisions
5,422,406
| Judgment layer | Sources | Assets | Promoted | Status | Promotion path |
|---|---|---|---|---|---|
| Explicit Knowledge | Policies, controls, procedures, accounting standards, treasury and credit rules | 457 | 29 | partial_runtime_promotion | Source document -> Policy Registry -> Policy Knowledge Graph -> Runtime rule -> evidence/replay gate |
| Control Intelligence | SOX controls, finance controls, operational controls, control metrics, risk registers | 13 | 13 | control_mapping_available | Control -> Policy -> Procedure -> Task -> Execution -> Evidence -> Attestation |
| Embedded Knowledge | SAP configuration, Ariba workflow, SOR contracts, approval hierarchy, live data wiring | 0 | 5 | source_contracts_available | System config -> semantic contract -> context object -> policy evaluation |
| Learned Knowledge | Historical decisions, exception outcomes, overrides, replay records, false positives | 5,422,406 | 5 | ledger_backed | Decision history -> pattern mining -> proposed rule -> steward approval -> policy runtime |
| Human Knowledge | SME guidance, controller judgment, playbooks, procedures, team practices | 18 | 0 | capture_required | SME capture -> structured judgment object -> review -> certification -> runtime use |
| Exception Intelligence | Audit findings, exceptions, escalations, non-standard approvals, manual workarounds | 36 | 0 | mining_required | Exception pattern -> root cause -> policy/control candidate -> governed promotion |
Policy Intelligence Layer
Policies stay in their systems of record. The runtime creates a governed intelligence layer that catalogs, structures, relates, evaluates, and replays policy usage.
Policy Source Layer
Policy Registry
Policy Knowledge Graph
Policy Runtime
Agent Runtime
Evidence + Replay + Attestation
Enterprise Judgment Layer
Operational judgment is broader than documents. It comes from policies, controls, systems, decisions, people, and history.
Formal Policies
Controls
Procedures
SAP / Workflow Configuration
Historical Decisions
SME Knowledge
Exception Handling
Audit Findings
Enterprise Judgment Layer
Credibility boundary
The platform should not replace enterprise policy repositories. It should operate as the Policy Intelligence and Enterprise Judgment Runtime that makes scattered policies, controls, procedures, system configuration, decisions, and SME knowledge executable, governable, explainable, and replayable.
Only canonical runtime policies should be described as enforceable today. Broader policy and judgment assets are candidates until they pass owner, version, evidence, control, and replay gates.
Enterprise positioning
The real asset is not the agent. It is the ability to capture, govern, explain, and continuously improve enterprise judgment across people, systems, controls, and decisions.