Three classes. Five traits. One schema.
Each source keeps its native shape. New entity types resolve at query time — no ETL program, no migration, no integration spend.
Daily settlement
Illustrative object. Five credentials, one owned business capability.
AI agents connect across all three — people, machines, and organizations and digital entities.
Humans are no longer the majority.
idGenius governs all three worlds in one context model. AI agents connect across every one of them.
Illustrative proportions. One schema, three worlds, one audit trail.
Scattered credentials become one owned capability.
Thousands of credentials, scattered across sources and named differently in each one.
Related identities converge into one owned group with a business meaning.
One governance object with an owner, a criticality and a lifecycle.
Illustrative. Owner, policy, lifecycle and audit trail attach to the bundle, not to each key.
Illustrative object. The same five traits attach to a human, a workload or a vendor.
2,000+
Years of ETL and millions of dollars — or read it where it already lives.
Every category. Every vendor.
idGenius integrates with the categories and vendors below for discovery — read in place, no migration, resolved into one schema.
One graph. Eleven entity classes. Live risk on every edge.
Triple store plus streaming ingest — the graph is the state, not a nightly snapshot.
Shared IdentityOne request. Deeply answered.
AuthZEN asks the question. idGenius fills the context that makes the answer meaningful.
Agent 123 — owned by Vendor A, acting for Project B, under Contract C, authorized by Employee D, accessing sensitive customer information, for a stated business purpose, under applicable policy — with current risk = X.
Every stage, exposed as an API.
Find every entity. Connectors and inventory, read in place.
Critical relationship mappings — what it is, what it reaches, what it risks.
Owner, policy and lifecycle bound to every object.
Decisions at your existing enforcement points.
Central intelligence. Maintains the graph, computes live risk, drives reaction.
Sponsored, approved, reached, expired — one trail per action.
Every decision traces to its evidence, with a cost-provenance record on every action.
One policy, published to every enforcement point.
Agent data-access policyNo proprietary SDK at the enforcement point — the contract is the integration.
Standard pipes. idGenius composes the context.
No proprietary SDK at the enforcement point — the contract is the integration.
Boring on purpose. Read at the bottom, decide at the top.
Five tiers. Every one of them a standard you already operate.
Nothing proprietary in the read path and nothing proprietary at the enforcement point. Swap any tier for the product you already own.
Every agent. Every action. One trail.
Engage to ReEngage. Governed by one ledger.
Humans run Joiner-Mover-Leaver. AI agents, machines and third parties don't fit that model.
Governance as a state, not a campaign. Every state change writes to the ledger — including the dormant years — so entitlements never need re-proving on a calendar: no quarterly access certifications, no manual re-attestation drives.
KuppingerCole's 2026 NHI Leadership Compass: JML-based tooling "was never designed to discover, attribute, or govern" machine and third-party identities at scale.
Where MCP and idGenius meet.
Model Context Protocol sits on the wire between models and tools. idGenius sits in the middle — so models ask in context.
MCP is the wire. idGenius is the lightweight AI gateway giving it meaning: every tool call is already routed, authorized and context-enriched — no separate gateway to deploy, license or maintain.
Policy isn't a PDF. It's a runtime.
Source documents become structured policy, and structured policy becomes live decisions on every request.
Documented obligations become a runtime all identities inherit. Every decision cites the clause it came from — so an auditor reads policy, not inference.
