Runtime system records
Maintain AI-system records with ownership, lifecycle stage, deployment target, endpoint references, and business-unit context in one governed workspace.
Feature · AI systems
Govern the runtime AI systems your organization actually deploys
SentinelAI extends governance beyond model records by giving teams a dedicated AI systems layer for the operational units that go live, inherit governed dependencies, and move through release and oversight decisions.
What this area covers
AI systems help teams govern what runs in production, not just the underlying components. Each record can connect models, approved use cases, datasets, readiness posture, and current release references so governance decisions reflect the deployed system boundary.
Related product areas
- Model registry
Maintain a governed inventory for AI models and use-case context with lifecycle state, ownership, risk posture, and supporting evidence.
- Prompt registry
Govern versioned prompts, retrieval settings, linked AI systems, and evaluation posture from a dedicated prompt operations record.
- RAG sources
Register governed retrieval sources with ingestion status, version history, citation context, and AI-system linkage.
- Evaluation suites
Define governed prompt evaluation suites with baselines, regression thresholds, run evidence, and release-blocking posture.
- Release governance
Manage AI-system release records with approval state, rollback references, dependency snapshots, and invalidation handling.
- LLM telemetry and monitoring
Bring live assurance signals, telemetry connector management, trigger rules, and evidence-ready monitoring context into AI governance workflows.
Core capabilities
Built to support production governance work
Linked governed dependencies
Roll up the models, use cases, and datasets that support the system so reviewers can see the operational dependency set without manual reconciliation.
Readiness posture
Track readiness states such as draft, ready, needs review, attention required, and retired so teams know which systems can progress and which need follow-up.
Release-aware oversight
Keep the current release reference and downstream release-governance state tied to the same runtime record used during review and monitoring.
Operational inventory clarity
Give governance, platform, and product teams a clearer picture of what is actually deployed instead of relying on model inventories alone.
Target users
- AI governance teams organizing runtime oversight around deployed systems instead of isolated assets
- Platform, release, and product owners who need one governed record for the operational unit going live
- Compliance and risk stakeholders reviewing readiness across the full system dependency set
- ML and data teams linking approved models and datasets into production-facing AI systems
Governance value
- Creates a governed runtime inventory that sits between model records and production release decisions
- Reduces ambiguity about which governed assets make up a deployed AI system
- Improves review readiness by keeping dependency context, ownership, and operational state together
- Supports release, evaluation, telemetry, and case workflows from a shared system boundary
- Makes post-deployment oversight easier to anchor to the system that actually changed
How teams use it
A practical operating flow for this feature family
Step 1
Define the system boundary
Register the runtime AI system with owners, deployment context, lifecycle state, and the release reference that matters operationally.
Step 2
Link governed dependencies
Connect the models, use cases, and datasets that should roll up into that runtime system before approval and monitoring work begins.
Step 3
Review and update over time
Use the AI-system record as the operational source of truth when readiness changes, releases move forward, or incidents require investigation.
Continue exploring
Explore how SentinelAI connects adjacent governance workflows
Model registry
Maintain a governed inventory for AI models and use-case context with lifecycle state, ownership, risk posture, and supporting evidence.
Prompt registry
Govern versioned prompts, retrieval settings, linked AI systems, and evaluation posture from a dedicated prompt operations record.
RAG sources
Register governed retrieval sources with ingestion status, version history, citation context, and AI-system linkage.
Evaluation suites
Define governed prompt evaluation suites with baselines, regression thresholds, run evidence, and release-blocking posture.
Release governance
Manage AI-system release records with approval state, rollback references, dependency snapshots, and invalidation handling.
LLM telemetry and monitoring
Bring live assurance signals, telemetry connector management, trigger rules, and evidence-ready monitoring context into AI governance workflows.