AI Fabrix operating model

An adoption map for architects and leaders — from enterprise reality to governed AI-assisted work through Role Assistants, with proof organized by business resources and capabilities in Evidence Fabrix (not chat threads).

Business value vs enabling pillars

Band	Pillars	Adoption question
Business value	Role Assistants, Evidence Fabrix	What outcomes and proof do we need?
Enabling	Enterprise Knowledge, Operational Trust	What data and controls make those outcomes safe?

Build and certify Connected Systems and Business Entities first. Role Assistants consume governed capabilities only after Operational Trust certifies scope.

Phase guide

Phase	Focus	Documentation
Discover	Product story and pillar bands	Understanding AI Fabrix
Design	Trust boundaries, roles, data scope	Operational Trust, Architecture
Integrate	Connected Systems → Business Entities → capabilities	Build AI-ready systems
Assure	Validate, upload, certify (three pillars)	Certification
Operate	Role Assistants, Content Review, evidence	Role Assistants, Evidence Fabrix, Operating model (cadence)

Evidence Fabrix — resource-based operating model

Evidence Fabrix answers audit and improvement questions in business terms:

Question	Evidence anchor
What resource was involved?	Business Entity + `resourceType` (deal, document, customer)
Which governed action ran?	Capability key (`deal:read`, `document:search`)
Who was accountable?	Role Assistant + active business role
What was the outcome?	Task completion signal — not conversation volume
What should we improve?	Capability evidence, role evidence, knowledge findings

%%{init: {
  "theme": "base",
  "themeVariables": {
    "fontFamily": "Poppins, Arial Rounded MT Bold, Arial, sans-serif",
    "fontSize": "16px",
    "background": "#FFFFFF",
    "primaryColor": "#F8FAFC",
    "primaryTextColor": "#0B0E15",
    "primaryBorderColor": "#E2E8F0",
    "lineColor": "#E2E8F0",
    "textColor": "#0B0E15",
    "borderRadius": 16
  },
  "flowchart": {
    "curve": "linear",
    "nodeSpacing": 34,
    "rankSpacing": 34,
    "padding": 10
  }
}}%%
%% diagram-type: connection-map

flowchart LR

classDef base fill:#FFFFFF,color:#0B0E15,stroke:#E2E8F0,stroke-width:1.5px;
classDef medium fill:#1E3A8A,color:#ffffff,stroke-width:0px;
classDef primary fill:#0062FF,color:#ffffff,stroke-width:0px;

BE["Business Entity<br/>(resource)"]:::medium
GC["Governed capability"]:::medium
RA["Role Assistant task"]:::medium
EV["Evidence Fabrix<br/>(proof by resource)"]:::primary

BE --> GC --> RA --> EV

Conversation history may help debugging; it is not the operational record. See Evidence vs conversation history.

Principles for rollout

Start with high-value roles and recurring tasks tied to certified resource types — not generic chat.
Require certification (operations, trust, governance) before broad Role Assistant adoption on a scope.
Treat governed capabilities as the only execution surface for AI — never raw vendor APIs.
Measure success with outcomes and resource-linked evidence, not message counts.
Keep humans in authority — Content Review, approval gates, and policy remain even as skill levels rise.
Expand scope only when Evidence Fabrix shows repeatable success on the current capability set.

Decision gates before expansion

Before a Role Assistant gains new capabilities or Business Entities:

Integration certified for that scope (three pillars + lifecycle)
Agent metadata and business semantics complete (verify-trust not notTrusted)
Governance sign-off on dimensions, protection, and Content Review rules where documents are involved
Operator runbook for failure, escalation, and evidence review

Skipping a gate moves risk from controlled platform behavior to ad hoc prompts and manual API use.

Who does what (steady state)

Role	Evidence Fabrix lens
Executive sponsor	Outcome metrics by capability and role — not chat analytics
Enterprise architect	Resource/capability taxonomy and certification rhythm
Integrator	Manifest truth for entities, capabilities, Content Review rules
Security / governance	Dimensions, protection, certification sign-off
Business operator	Task outcomes, steward queues, escalation
Platform operator	Re-certify triggers after vendor or policy change

Organizational cadence detail: Operating model.

Limits

Skill levels, promotion rules, and impact dashboards depend on your AI Fabrix release and tenant configuration. This page describes the adoption model — not your tenant's certification profile or live Role Assistant catalog.

Documentation Index