Overview

Governe: central and per-org governance for a sovereign agentic platform. SSO, roles, model access, audit, quotas, and retention.

Governe is what makes Prisme.ai an Information System for AI rather than a pile of assistants, the console where governance runs on two levels at once. A central governance sets platform-wide rules; each organization inherits from it and adapts within those guardrails. The result is the balance enterprises need: freedom and agility for teams, compliance guaranteed across the board.

From here the platform handles the cross-cutting concerns once, for everyone: SSO and centralized permissions, audit logs for full traceability, AI model access control, and chargeback / rebilling so usage and cost are attributed per organization. It serves both organization administrators and platform super-admins.

Video Tour: Governance

This video is currently available in French.

Key Concepts

Organizations

An Organization is the primary tenant unit in Prisme.ai. Each organization contains:

• Members with assigned roles
• Groups for bulk member management
• Roles defining permissions
• API Keys for programmatic access
• Service Accounts for machine-to-machine authentication
• Subscriptions with usage quotas

Users can belong to multiple organizations and switch between them from the platform header.

Two Administration Levels

Governe serves two distinct user groups:

Level	Scope	Typical Users
Organization Admin	Single organization	IT admins, team leads
Platform Admin	All organizations	Platform operators, super-admins

Features at a Glance

Identity & Access

Members, groups, roles, permissions, SSO, and join rules

Model Governance

Control which AI models are available, set usage policies, and manage per-agent controls, knowledge storage, and data retention

Capabilities Catalog

Manage MCP servers, tools, guardrails, and skills

Observability

Org-wide platform operations and LLM costs; per-workspace deep-dive in Builder

API Keys & Service Accounts

Programmatic access with scoped permissions

Quotas & Subscriptions

Usage limits, rate controls, and billing tiers

Branding & Navigation

Customize appearance, logos, and menu structure

Audit Logs

Track all administrative actions for compliance

Navigation Structure

When you open Governe, the sidebar shows organization sections:

Section	Description
Overview	Organization dashboard with stats and quick actions
Agents Controls	Model governance, usage, service accounts, per-agent controls, knowledge storage, and data retention
Capabilities	MCP servers and tools catalog
Observability	Org-wide platform operations and LLM costs (per-workspace deep-dive now in Builder)
Members	User management and invitations
Groups	Organize members into groups
Roles	Define and assign permission sets
Invites	Generate invitation codes
Join Rules	Auto-assign roles based on user attributes
SSO	Configure identity providers (OIDC/SAML)
API Keys	Manage programmatic access credentials
Quotas	View usage against subscription limits
Subscription	Billing tier and feature access
Audit	Activity logs for compliance
Appearance	Branding and visual customization
Menu Editor	Customize navigation structure

Platform admins see additional sections for managing all organizations, users, models, and subscriptions.

Getting Started

1. Access Governe

   Navigate to Apps > Governance from the platform sidebar, or go directly to /apps/ai-governance.

2. Select Your Organization

   If you belong to multiple organizations, select the one you want to manage from the dropdown.

3. Review the Dashboard

   The Overview page shows member count, role count, API keys, and usage quotas at a glance.

4. Configure Access

   Start by setting up roles, inviting members, and configuring SSO if needed.

Who Uses Governe?

IT Administrators

• Manage organization members and access
• Configure SSO with corporate identity providers
• Create API keys for integrations
• Monitor usage and enforce quotas

AI Operations

• Control which AI models are available
• Set model routing and failover policies
• Monitor LLM costs and token usage
• Configure guardrails and capabilities

Security Teams

• Implement role-based access control
• Review audit logs for compliance
• Manage service accounts for automation
• Configure join rules for automatic provisioning

Platform Admins

• Manage all organizations across the platform
• Configure available AI models globally
• Create subscription tiers with quotas
• Monitor platform-wide health and costs

Next Steps

Identity & Access

Members, roles, groups, SSO, and permissions

Model Governance

Control AI model access and usage policies

Capabilities

MCP servers, tools, and guardrails catalog

Observability

Monitoring, metrics, and request tracing