Skip to main content
IIOS — The platform for organizational reasoning
Platform Architecture

The reference architecture for organizational reasoning.

A technical view of IIOS for CIOs, security leaders, investors, and integration partners — how data flows from your existing systems into a governed, explainable reasoning environment.

Design Principles

Architectural commitments that hold at every layer

These principles are not features bolted on later. They shape how every layer of IIOS is designed.

A reasoning layer, not another system of record

IIOS sits above your existing stack and reasons across it. It does not replace your ERP, CRM, PLM, or BI — it makes them more valuable.

Customer-owned data by default

The platform reads and reasons over data you continue to control, minimizing replication rather than creating a duplicate enterprise data store.

Explainable by construction

Every assertion carries source provenance and confidence. Conclusions can be traced from question to evidence to answer.

Deploy where policy requires

The same architecture runs as managed SaaS, in your cloud tenancy, dedicated, or fully air-gapped — without changing the reasoning model.

The Stack

From systems of record to institutional judgment

Each layer has a distinct responsibility. Together they turn disconnected systems into a shared reasoning environment.

Systems of Record
ERPCRMPLMData warehouseDocuments
Governed Connectors & Ingestion
Read-only APIsReplicas & viewsSecure file exchange
Knowledge Fabric
NormalizationEntity resolutionProvenance
Industrial Graph & Ontology
EntitiesRelationshipsEvents
Reasoning Engine
InferenceEvidenceConfidence
Studios & Applications
Role lensesWorkflowsAPIs & SDK
People & Decisions
Data Flow

How a question becomes a defensible answer

Reasoning is a traceable path through the graph — not a black box.

01

A question enters

A person or Studio poses a question in natural language or through a structured workflow.

02

The graph is traversed

The reasoning engine traverses entities and relationships in the Industrial Graph rather than matching keywords.

03

Evidence is assembled

Relevant sources, prior decisions, and institutional memory are gathered, each with provenance and confidence.

04

An answer is reasoned

The platform proposes a conclusion with its supporting evidence. People decide; the decision is remembered.

Deployment Topology

One architecture, four deployment models

The reasoning model is identical across deployments. What changes is where it runs and who holds the keys.

Shared SaaS

Managed multi-tenant with logical isolation. Fastest path to value.

Customer Cloud

Runs in your AWS, Azure, or GCP account. Data and compute stay in your tenancy.

Dedicated Enterprise

Single-tenant, isolated infrastructure with customer-managed keys.

On-Prem / Air-Gapped

Runs inside controlled, disconnected, or classified environments.

For the full security posture, governance model, and AI oversight controls, see Security & Data Control.

Integration Surface

Built to connect to what you already run

IIOS reasons across the enterprise stack through governed, auditable connections.

ERPCRMPLMSharePointSnowflakeSQL databasesMarket data providersM&A databasesDocument repositories

An extensibility surface for technical partners

Studios and applications are built on the same APIs and SDK available to your teams and partners. The graph is queryable, the reasoning is composable, and new Studios extend the platform without forking it.

Bring your architecture and security teams.

Request the full reference architecture, deployment guide, and governance documentation for your technical evaluation.