Development may own code, infrastructure may own cloud resources, QA may provide test evidence and DevOps may maintain pipelines. The hand-offs still leave a gap if nobody is accountable for how the service behaves after release.
Trace one service end to end
Choose a customer or operational service and follow a change from approval to production. At each step, identify the decision owner, the evidence required and the route for escalation. Repeat the exercise for one recent incident, from first signal through recovery and prevention work.
Questions that reveal a gap
- Who can stop a high-risk release, and on what evidence?
- Who owns rollback and recovery decisions?
- Which alerts represent customer or business impact?
- Who leads an incident and keeps stakeholders informed?
- Who verifies that prevention actions are completed?
If different teams give incompatible answers, document the disagreement before buying another tool. CI/CD and observability can support an operating model; they cannot decide ownership or escalation on their own.
A simple ownership record
For each critical service, record the accountable owner, release approver, incident lead, recovery authority, key dependencies and review date. The same person need not perform every task. The aim is to make decision rights and hand-offs explicit.
After an incident, review impact, detection, escalation, recovery and the conditions that allowed the failure. Assign a prevention action to an owner and revisit whether the change reduced the chance or impact of recurrence.