Manufacturing / Commercial vehicles
Daimler Truck: Investigating incidents and preparing code-fix pull requests for an aftersales portal
- Company
- Daimler Truck
- Country
- Not classified
- Adoption stage
- Limited operation
- Source published
- Date basis
- The date the source was published. It can differ from the date adoption started.
- How the source was checked
- Read the full source text
The work problem
When incorrect settings or memory limits disrupted Daimler Truck's TASP! aftersales portal, operations engineers had to read alerts, investigate causes and prepare fixes. The AWS story says this manual process took an average of 2 to 3 hours per incident. Repeated incident work occupied engineering capacity and required separate documentation of investigations and remediation rationale.
Technology and data
Daimler Truck's IT Aftersales team worked with CGI to build a multi-agent workflow using Bedrock AgentCore, AWS DevOps Agent and Lambda. Monitoring alerts lead to an investigation summary and proposed diff, and fixes are applied in a sandbox only for allowed repositories. After checks pass, the system opens a GitHub pull request containing the incident timeline, root cause, proposed fix and test results. Engineers review and approve before merging. Alert deduplication and a cost circuit breaker provide safeguards, and the scope is restricted to thoroughly tested incident types.
Results
AWS reports that the workflow, including automation and human review and approval, takes 15 minutes, reducing resolution time by approximately 88%. Automatically generated pull requests preserve timelines and remediation rationale, supporting documentation and audit preparation. This is not a claim of autonomous resolution of every incident; support for more incident types remains a future plan.
Limits and open questions
The article describes existing sandbox checks but also lists additional testing as future work, leaving the precise current testing boundary unclear. The number of measured incidents and measurement period were not disclosed.
Sources
- Automating incident resolution with a multi-agent self-healing SRE solution on AWSAWS for Industries, Accessed
Compiled from public sources. These are not results from ATF Works customers.