Platform Overview

This section documents how Modulos works in practice: the core objects, workflows, permissions, and integrations you’ll use to run AI governance at scale. Each subsection below covers one major module of the platform.

Organizations

Workspace boundary, membership, settings, and shared libraries

Projects

The unit of governance scope for an AI system or initiative

Governance

Frameworks, Requirements, Controls, Evidence, and reviews

Risk

Organization taxonomy, project risks, quantification, and risk limits

Runtime Inspection

Sources, tests, schedules, results, and remediation workflows

Integrations

API tokens, Scout Connectors, and external data access

How the platform is organized

Modulos uses a small set of core objects and connects them with explicit relationships so that every governance action is traceable from Requirement to Evidence.

• Organizations are the top-level workspace. They define membership, roles, global defaults (language, currency, risk taxonomy), and feature availability. Every user belongs to exactly one organization.
• Projects represent a concrete governance scope — typically one AI system, product, or initiative. Projects contain frameworks, Controls, Evidence, risks, and test results. They are the primary unit of work in Modulos.
• Governance is the compliance audit trail. Frameworks define Requirements, Controls implement them, Evidence proves them, and reviews make every status change auditable. The governance module supports multi-framework mapping so a single Control can address Requirements from the EU AI Act, ISO 42001, and NIST AI RMF simultaneously.
• Risk provides quantitative risk management. Organizations define a risk taxonomy with categories and impact scales, then project teams assess individual risks, quantify them in monetary terms, and track exposure against risk limits.
• Runtime Inspection provides continuous validation. Connect data Sources, define tests against governance-relevant properties (bias, performance, security), schedule them to run on a cadence, and link results back to Controls as automated Evidence.
• Integrations provide authenticated access to external systems. Project-level Sources connect monitoring and context systems (Prometheus, Datadog, OpenTelemetry, Modulos Client, GitHub, Azure, Langfuse, AWS, Vijil, Snowflake, Claude Console). User-level Scout Connectors (Google Drive, GitHub, Atlassian, Bitbucket) let the AI assistant query connected sources during chat.

Availability

Depending on your organization setup and subscription, you may not see every module in the navigation.

If a page referenced in the docs is missing in your workspace, ask an organization admin to confirm access.