The Use Case

Estimated time: 10–15 minutes

Acme Inc. is facing issues within their CI/CD system, and they’re asking you to build a Gen AI based solution for them.

The Challenge

Acme Inc. has a critical CI/CD pipeline that frequently fails due to various issues such as:

  • Build failures due to dependency conflicts

  • Deployment errors caused by configuration mismatches

  • Resource constraints leading to timeouts

  • Integration test failures requiring manual investigation

The current process requires developers to:

  • Manually investigate error logs

  • Search through documentation for solutions

  • Create GitHub issues for tracking problems

  • Manually restart failed pipelines

  • Coordinate with team members for resolution

This manual intervention is:

  • Time-consuming and delays releases

  • Error-prone due to human oversight

  • Not scalable as the team grows

  • Lacking consistent documentation of solutions

How to think about the problem (DORA and MTTR)

  • DORA metrics are industry-standard measures of software delivery performance used to identify high performing teams: Deployment Frequency, Lead Time for Changes, Change Failure Rate, and Mean Time to Recovery (MTTR). See https://dora.dev for background.

  • MTTR is the average time it takes to restore service after a failure. Lower MTTR means faster recovery and less downtime.

  • In later modules you will codify the agent and wire it into the pipeline’s finally step so failures automatically trigger the agent—helping drive down MTTR.

The Solution: AI-Powered CI/CD Agent

We will build an intelligent agent that can:

  • Detect and analyze CI/CD failures

  • Interact with OpenShift to query cluster resources and pod status

  • Search for relevant solutions using web search and documentation

  • Provide actionable recommendations to developers

  • Create GitHub issue in the specified repository for tracking