Support sees three tickets about failed exports in one hour. Use Driver to compare their evidence and prepare a focused question for on-call while your team runs its incident process.
Gather the incident inputs
Give Driver the Jira ticket URLs, service and environment, error signatures, account identifiers, and the first-failure timestamps with a timezone. Paste relevant, timestamped deployment records and log excerpts from your incident tools, with sensitive details removed. Ask the release owner which codebase branches or supplied diffs correspond to the affected deployments.
Compare the reports and involve on-call
- Start an investigation with the ticket URLs and inputs. Ask: “Compare these three export failures. Make one row per ticket with account, environment, first failure, error signature, and source evidence. Which reports have evidence for a common cause, and which remain separate?”
- Ask Driver to trace the matching code paths and compare the supplied deployment timeline. For each candidate cause, request the observation that would rule it in or out.
- Copy the comparison and open questions into the incident issue. Ask on-call which cause is to blame and to run the proposed check. For example, distinguish an export timeout from a permission failure even when both customers reported a failed export.
- Each responder starts an independent Driver conversation with the incident issue URL and the latest evidence, or pastes the reviewed comparison. Put new findings on that issue so responders can carry them into their own investigations.
Confirm recovery and update customers
After the team deploys a fix or rolls back, give Driver the deployment record and the results of your operational checks. Ask it to summarize which ticket symptoms recovered and what remains unresolved. Your team runs the checks in the incident environment; Driver compares the supplied evidence.
Ask for a customer update in the conversation, review the affected window and recovery claims, and copy the result to the relevant Jira tickets. Include any customer action, such as retrying an export. A common symptom becomes a common-cause finding only when the evidence connects the reports.