Skip to main content

Governance

Enterprise-grade governance framework for enforcing compliance, security, and quality standards across all Strongly AI platform resources. Define policies, track compliance through solutions and snapshots, configure AI guardrails, control resource promotion between environments, and maintain a complete audit trail for regulatory requirements.

Core Capabilities

CapabilityDescription
Policy ManagementDefine, version, and enforce compliance policies with custom stages, validators, and enforcement rules
Policy CatalogShare and discover reusable policy templates with ratings and reviews
Stages & Approval WorkflowsMulti-stage approval workflows with user, role, and group validators
SolutionsGroup related platform components for unified governance and compliance tracking
Compliance SnapshotsCapture point-in-time compliance state for audit evidence, releases, and regulatory review
AttestationsDigital compliance statements with expiration tracking and revocation support
AI GuardrailsContent filtering, PII detection, prompt injection prevention, and rate limiting for AI models
Enforcement EnginePre-deployment compliance checks with hard blocks, soft blocks, warnings, and override management
Environment PromotionControlled promotion workflow with policy compliance gating
Audit TrailImmutable record of every governance action with full before/after state capture
NotificationsIn-app notifications for approval requests, decisions, escalations, and compliance alerts

Governable Resource Types

Governance policies can target the following platform resource types:

Resource TypeKeyGuardrailsPromotionDescription
Applicationsapp-YesDeployed marketplace and custom applications
Add-onsaddon-YesManaged database and infrastructure services
Workflowsworkflow-YesAutomation and AI agent workflows
Data SourcesdataSource--External database and API connections
ML ModelsmlModel-YesML Registry models and experiments
AI Gateway ModelsaiGatewayModelYes-LLM and generative AI models
Workspacesworkspace--Collaborative team workspaces
Projectsproject--Project containers for resources

Each resource type has specific policy fields. For example, AI Gateway Models support guardrail requirements, content filtering levels, rate limits, and cost budgets. Applications support security scan requirements, code review, and deployment approval. See Enforcement for details on resource-specific policy fields.

Getting Started

1. Create Your First Policy

Navigate to Governance in the sidebar and click Create Policy. The Policy Builder wizard guides you through defining the policy name, category, severity, stages, validators, and enforcement rules. See Policies for step-by-step instructions.

2. Apply a Policy to a Resource

  1. Open any resource detail page (Application, Workflow, AI Model, etc.)
  2. Click the Governance tab
  3. Click Apply Policy
  4. Select a policy from the dropdown (filtered to applicable resource types)
  5. A policy instance is created with status not_started

3. Complete Policy Stages

  1. Open the policy instance from the Governance > Validation page
  2. Complete each stage sequentially:
    • Fill out all required fields
    • Upload any required artifacts or evidence
    • Submit the stage for validator approval
  3. Assigned validators receive notifications and can:
    • Approve — Stage completes, workflow advances to the next stage
    • Deny — Stage is rejected, the requestor must revise and resubmit
    • Conditional Approval — Approve with documented conditions
  4. Once all stages are approved, the policy instance status becomes completed

4. Group Resources into Solutions

Solutions group related platform components (apps, workflows, data sources, models) for unified compliance tracking. Navigate to Governance > Solutions and click Create Solution to define a solution with explicit component selection, tag-based selection, or query-based selection. See Compliance for details.

5. Take Compliance Snapshots

Snapshots capture a point-in-time view of a solution's compliance state. Create snapshots for baseline assessments, periodic reviews, release approvals, or formal audits. Frozen snapshots serve as immutable audit evidence. See Compliance.

6. Configure AI Guardrails

For AI Gateway models, configure guardrails to enforce content filtering, PII detection, prompt injection prevention, rate limiting, and cost controls. Guardrails are applied on every API request to the model. See Guardrails for the complete configuration guide.

7. Monitor Compliance

The Governance Dashboard provides real-time metrics:

MetricDescription
Active PoliciesNumber of enforced policy definitions
Compliance RatePercentage of policy instances that are completed
Pending ApprovalsInstances awaiting validator action
Total ResourcesResources currently under governance

Quick actions on the dashboard provide one-click access to create policies, create solutions, create snapshots, review pending approvals, and view compliance reports.

Access Control

Governance enforces strict access control at every level:

ActionWho Can Do It
Create policiesAny authenticated user
Modify/delete policiesPolicy creator or administrator
Share policiesPolicy owner or administrator
Apply policy to resourceUsers with resource access
Submit stage for approvalInstance creator or administrator
Approve/deny stagesAuthorized validators (user, role, or group) or administrator
Create deployment overridesAdministrators only
Delete completed instancesNot allowed (audit trail preservation)

Validator Types

Stage validators can be assigned by three mechanisms:

TypeFormatBest For
UserSpecific user IDDesignated reviewers, compliance officers
RoleRole name (e.g., admin)Flexible team assignments
Grouporg:<orgId>:role:<role> or organization IDCross-team approvals, committee reviews

Compliance Frameworks

Tag policies and templates with industry compliance frameworks for easy discovery and mapping:

  • SOC 2 — Security, availability, and confidentiality controls
  • HIPAA — Healthcare data privacy and security
  • GDPR — EU data protection requirements
  • ISO 27001 — Information security management
  • ISO 42001 — AI management system standard
  • PCI DSS — Payment card industry standards
  • NIST AI RMF — AI risk management framework
  • EU AI Act — European AI regulation

Architecture Overview

┌─────────────────────────────────────────────────────────────┐
│ Governance Dashboard │
│ Metrics · Quick Actions · Policy Distribution · Activity │
├──────────┬──────────┬──────────┬──────────┬────────────────┤
│ Policies │Solutions │Snapshots │Guardrails│ Enforcement │
│ │ │ │ │ │
│ Define │ Group │ Capture │ AI model │ Pre-deployment │
│ stages & │ platform │ point-in │ content │ compliance │
│ rules │ resources│ -time │ safety │ checks │
│ │ │ state │ controls │ │
├──────────┴──────────┴──────────┴──────────┴────────────────┤
│ Enforcement Engine │
│ hard_block · soft_block · warning · override management │
├─────────────────────────────────────────────────────────────┤
│ Audit Trail │
│ Every action logged · Before/after state · Immutable │
├─────────────────────────────────────────────────────────────┤
│ Notification System │
│ Approvals · Escalations · Deadlines · Attestation expiry │
└─────────────────────────────────────────────────────────────┘

All governance features are accessible from the Governance section in the platform sidebar:

PagePathDescription
Dashboard/governanceOverview metrics and quick actions
Policies/governance/policiesBrowse and manage policy definitions
Create Policy/governance/policies/newPolicy Builder wizard
Policy Catalog/governance/catalogBrowse and use policy templates
Solutions/governance/solutionsManage solutions
Create Solution/governance/solutions/newSolution Builder
Snapshots/governance/snapshotsBrowse compliance snapshots
Create Snapshot/governance/snapshots/createSnapshot creation wizard
Compliance Monitor/governance/complianceReal-time compliance across all resources
Validation/governance/validationReview and approve pending instances
Promotions/governance/promotionsEnvironment promotion control
Audit Log/governance/auditComplete audit trail viewer
Best Practice

Apply policies early in the development lifecycle. Use warning enforcement in Development, soft_block in Staging, and hard_block in Production to catch issues progressively before they reach production.