Release records for AI systems
Anchor each rollout to an AI system with a governed version, release name, target environment, and descriptive metadata.
Feature · Release governance
Govern AI-system promotions with release records built for oversight
SentinelAI gives teams a governed release timeline for AI systems, tying each rollout to approval state, dependency snapshots, rollback references, and evaluation-aware promotion controls.
What this area covers
Release governance makes AI deployment decisions visible and durable. Teams can create release records for runtime AI systems, route them through approval and promotion steps, preserve rollback pointers, and invalidate records automatically when prompts or retrieval dependencies change.
Related product areas
- AI systems
Track governed runtime systems that combine models, approved use cases, datasets, release state, and readiness into one operational record.
- 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.
- Governance cases
Coordinate alerts, findings, remediation, evidence posture, SLA deadlines, and closure outcomes in one shared case workspace.
- Compliance workflows
Operationalize evidence collection, control tracking, remediation, and framework mapping across AI systems.
Core capabilities
Built to support production governance work
Approval and promotion workflow
Move records through draft, pending approval, approved, released, rejected, and revoked states with clear workflow transitions.
Dependency snapshots
Preserve the linked prompt and RAG-source set associated with the release so reviewers know what was part of the promotion decision.
Rollback references
Record rollback pointers and historical release relationships so teams can investigate and recover from release issues faster.
Dependency invalidation
Mark releases for re-review when linked prompts or sources change, making it harder for approvals to drift away from the current dependency state.
Target users
- Release and platform owners coordinating governed AI-system promotions
- AI governance teams that need deployment state tied to approval and dependency history
- Prompt, retrieval, and ML teams whose changes can invalidate a prior release decision
- Risk and assurance stakeholders reviewing whether release evidence is still current
Governance value
- Makes AI release approvals reviewable instead of leaving them inside developer tooling alone
- Improves traceability between runtime promotion decisions and the dependencies they relied on
- Connects evaluation posture and rollback planning to the same governed release record
- Surfaces when a once-approved release needs re-review because supporting assets changed
- Helps teams explain release readiness to governance councils and auditors with less reconstruction
How teams use it
A practical operating flow for this feature family
Step 1
Create a governed release record
Attach the rollout to the AI system, capture the target environment, and preserve the current prompt and source dependency set.
Step 2
Run approval and promotion steps
Move the record through review, approval, promotion, rejection, or revocation with the surrounding governance context attached.
Step 3
Respond to dependency change
Invalidate or re-review release records when prompts, sources, or evaluation posture change after approval.
Continue exploring
Explore how SentinelAI connects adjacent governance workflows
AI systems
Track governed runtime systems that combine models, approved use cases, datasets, release state, and readiness into one operational record.
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.
Governance cases
Coordinate alerts, findings, remediation, evidence posture, SLA deadlines, and closure outcomes in one shared case workspace.
Compliance workflows
Operationalize evidence collection, control tracking, remediation, and framework mapping across AI systems.