Skip to content
.md

Regulated Industries

In regulated industries—finance, healthcare, government, critical infrastructure—every code change needs accountability. Who approved it? What tests passed? When did it merge? A merge queue provides this by design.

Auditors want to know:

  • Who approved each change
  • What was tested before deployment
  • When changes landed in production
  • Why a change bypassed normal process (if ever)

Manual processes create gaps. Someone merges directly. Approvals happen in Slack. Tests get skipped “just this once.” These gaps become audit findings.

A merge queue enforces process automatically:

RequirementWithout QueueWith Queue
All changes reviewedHope and policyEnforced gate
CI passes before mergeUsuallyAlways
Audit trailScattered across toolsSingle source of truth
No direct pushesHonor systemTechnically impossible

The queue becomes your compliance control—not a document someone might ignore.

Every PR that enters the queue is logged:

  • Entry time
  • Required approvals (who, when)
  • CI results (which checks, pass/fail)
  • Merge time
  • Final commit SHA

This trail is automatic. No one needs to remember to document.

Configure required approvals:

  • Minimum reviewers (e.g., 2 approvals)
  • Specific team sign-off (security, compliance)
  • No self-approval
  • Fresh approval after changes

The queue won’t process a PR until gates pass.

Every change runs the same checks:

  • No skipping tests for “urgent” fixes
  • No merging with failing CI
  • Consistent validation across all changes

Auditors love consistency.

Link every deployment to:

  • The PR that introduced it
  • The approvers who signed off
  • The test results that validated it
  • The ticket/issue it addresses

When an auditor asks “how did this code get to production?”, you have the answer.

SOC 2 requires controls around change management. A merge queue provides:

SOC 2 RequirementQueue Feature
Changes are authorizedRequired approvals
Changes are testedMandatory CI
Changes are documentedAudit trail
Access is controlledNo direct push to main

For payment systems, PCI-DSS demands:

  • Separation of duties → Different people approve and merge
  • Code review → Required reviewers before queue entry
  • Testing → CI must pass before merge
  • Change records → Complete audit trail

Healthcare systems need:

  • Access controls → Only authorized users can approve
  • Audit logs → Full history of who did what
  • Integrity controls → CI validates no unauthorized changes

Financial reporting systems require:

  • Change authorization → Approval gates
  • Testing evidence → CI results as proof
  • Segregation of duties → Reviewer ≠ author enforcement

Regulations don’t disappear for emergencies—but process can adapt.

Use priority management for urgent fixes:

  • Hotfix jumps the queue
  • Still requires approval (maybe fewer reviewers)
  • Still runs CI (maybe a faster subset)
  • Fully logged as expedited

If something truly bypasses process:

  • Log it explicitly
  • Document the justification
  • Review in post-incident
  • Track as an exception

A queue makes exceptions visible rather than hidden.

Link PRs to tickets (Jira, ServiceNow):

  • Require ticket reference in PR
  • Auto-update ticket on merge
  • Trace deployments to change requests

Export audit data to:

  • GRC platforms (ServiceNow GRC, Archer)
  • SIEM systems (Splunk, Datadog)
  • Custom compliance dashboards

Integrate with existing approval systems:

  • Require compliance team sign-off for sensitive paths
  • Auto-request security review for specific file changes
  • Block merge until external approval completes

Generate reports showing:

  • All changes in a time period
  • Who approved each change
  • CI pass rate
  • Policy exceptions and justifications
  • Time from approval to production

When auditors visit, you export a report instead of scrambling.

  1. Automation beats documentation — Enforced controls are stronger than written policies
  2. Consistency matters — Every change through the same process
  3. Trails are automatic — No manual logging required
  4. Exceptions are visible — Priority changes are tracked, not hidden
  5. Auditors love queues — Single source of truth for change management