Change Control and Rollback
Version, diff, and roll back any platform resource
Saga is built-in version control and change management for every configurable resource on the platform. Each change is recorded with who made it, a comparison against earlier versions and a one-step rollback, and drafts stay separate from the live version that systems actually run. Where you choose it, an approval gate sits in front of production, giving IT and security teams an audit-grade record of every configuration change.
Service desks govern tickets about a change; Saga governs the change itself. Every configurable resource on the platform carries its own version history, diff, and one-step rollback, drafts stay apart from what production runs, and approval workflows sit in front of production where you choose. Standing AI reviewers flag risks, while a person always casts the binding vote.
- Version history, diff, and rollback on every configurable resource
- Conflict detection, so two editors never silently overwrite each other
- Approval workflows with change windows, risk categories, and reviewers
- Full audit trail covering changes, conflicts, and post-implementation review
- Standing reviewer agents flag risks and recommend, while a human casts the binding vote
What you get
What Saga gives your team.
History, diff and rollback
A timeline and side-by-side comparison of any two versions on every versioned resource. Rollback creates a new version, so history is never destroyed.
Conflict detection and drafts
A save against a stale copy returns a conflict to review. Drafts and live versions stay separate, and running systems use the approved version.
Change requests and approvals
Requests carry categories and risk levels, route to named approver lists, and accountability rules set stricter approval counts for high-value resources.
Drift detection
Re-verifies a resource's live configuration against its recorded history and captures an out-of-band change as its own attributed version.
Scheduled changes and freezes
Schedule a change to apply at a chosen time, with change windows and freeze periods driven by named schedules and collisions shown beforehand.
AI-assisted review
A plain-English summary and risk assessment on each request, plus standing reviewer agents whose recommendations sit beside human approvers.
How it works
- 1
Edit as usual; history builds itself
Any configurable resource records per-component versions with a content hash, and repeated autosaves squash into one rolling entry. A save against a stale copy returns a conflict to review.
- 2
Turn on change management
Opt in per product and organization. Require change requests with categories and risk levels, named approver lists, and stricter approval counts for high-value resources.
- 3
Review and approve requests
Approvers work from an inbox showing only what they can decide, with wait time and service-level flags. AI summaries and risk assessments sit beside the human vote, which always decides.
- 4
Promote, monitor and roll back
Approved changes are promoted; running systems use the approved version, not the draft. A guarded rollout offers a one-click human-confirmed rollback on an error spike, and rollback creates a new version.
Ask AI about Saga
Every Saga page has an Ask AI button. It opens a chat that already has the page, its rows, and your selection. Try asking:
- Which change requests on this page have waited longest for approval?
- Which pending requests carry the highest risk level?
- Which change templates here cover recurring standard changes?
How teams use it
- Require a second approver before a production access policy changes.
- Roll back an AI routing change that raised error rates, after a person confirms.
- Show an auditor who changed what, who approved it and when.