Legacy Technology · Proof
Doubling internal incident detection for an automotive software provider
Our site reliability engineers helped a North American provider of dealership software find more of its own incidents before customers reported them.
At a glance
| Client | A North American automotive dealership software provider, engaged through a partner |
|---|---|
| Service | Site reliability engineering consulting |
| Headline result | Internal incident detection rate rose from 35% to 70% |
| Scale of analysis | 444 Azure DevOps work items analysed (to February / March 2026) |
| Reporting | Quarterly management reports, with the source data linked from every chart |
The challenge
Too many incidents were being found by customers before the client’s own monitoring caught them. Management needed to see where incidents came from and whether reliability work was paying off.
What we did
- Analysed incident and work-item history to find where detection was failing.
- Worked with the client’s teams to improve monitoring and alerting where gaps were found.
- Introduced quarterly reporting for management, with every chart linked to its source data so the numbers can be checked.
Results
The share of incidents detected internally, rather than reported by customers, rose from 35% to 70%. Management now reviews reliability each quarter from reports they can trace back to the raw data.
Why this matters for legacy systems
Old systems rarely have good monitoring. The same discipline (detect first, write it down, report honestly) is what our legacy support pods bring to C, Pro*C and Unix estates.
Tell us what you’re running
Send us the stack and the problem. We’ll reply within one working day and, if it’s a staffing request, send a shortlist within five.