In today’s world, it’s almost unimaginable to develop software without CI/CD (continuous integration/continuous delivery) pipelines. Automated build and test processes ensure consistent quality. However, when a pipeline fails, DevOps teams often must undertake a time-consuming process of searching log files for error patterns, identifying causes, and creating tickets.
A project in which our colleague Yu Zou from sepp.med participated shows that AI can drastically reduce this effort.
Anyone working on DevOps projects is familiar with this scenario: The pipeline turns red, and the detective work begins. Teams open log files containing hundreds or thousands of lines, search through them manually to find the cause of the error, and then create a ticket in the project board by hand.
What often falls by the wayside:
This manual approach is error-prone and unsustainable. The larger the project, the more pipeline runs and logs there are, and the longer searches take.
The consequences extend beyond mere lost time.
This is particularly serious in safety-critical industries, such as medical technology. A lack of systematic risk analysis of build processes can create gaps that are relevant to audits and jeopardize traceability.
During an internal AI hackathon at a large medical technology company, Yu Zou and two colleagues solved this problem on the spot. In just three days, they developed the Pipeline Oracle tool, which won an award in the “Learning Focused Solution” category.
The principle: Pipeline Oracle connects to the Azure DevOps server via API, retrieves logs from failed pipelines, and sends them to a large language model (LLM). The AI analyzes the errors, categorizes them, suggests specific solutions with confidence values, and automatically creates a risk analysis. If desired, the tool can also create issues in the project board and assign them to the responsible person. All results are summarized in a clear dashboard.
What makes the tool so valuable is its change in perspective. Rather than reactive troubleshooting, Pipeline Oracle enables proactive data analysis. It recognizes error patterns across multiple runs, predicts future problem areas, and evaluates release readiness.
It was developed entirely in Python and is universally applicable.
This hackathon project demonstrates how AI can be integrated into existing development processes with a clear focus on a specific use case, pragmatic implementation, and measurable benefits.
This approach is precisely what underlies the AI ABC Framework, which we use to support companies in their structured introduction to AI. The key is to find the right starting point and achieve initial results quickly, whether the focus is on DevOps automation, quality assurance, or documentation.
To learn more about Pipeline Oracle, attend the Afterwork Exchange on March 19, 2026, where Yu Zou will present the tool.
Pipeline Oracle is an AI-powered analysis tool for CI/CD pipelines. It automatically reads log files, categorizes errors, suggests solutions, and generates risk and forecast reports on a centralized dashboard.
Although Pipeline Oracle was developed for Azure DevOps, its configurable structure makes it transferable to other CI/CD environments.
The hackathon demonstrates how AI can generate tangible value in software development processes. As part of our AI consulting, we identify and implement precisely such use cases.
With our AI Readiness Check, we analyze your current situation, identify appropriate use cases, and develop a manageable roadmap for your successful entry into the world of AI.
Firstname:
Lastname:
E-Mail Address:
Phone:
Subject:
Your message:
Yes, I consent to my personal data being collected and stored electronically. My data will only be used for the purpose of responding to my inquiry. I have taken note of the privacy policy.
You are currently viewing a placeholder content from OpenStreetMap. To access the actual content, click the button below. Please note that doing so will share data with third-party providers.
You need to load content from hCaptcha to submit the form. Please note that doing so will share data with third-party providers.
You are currently viewing a placeholder content from Google Maps. To access the actual content, click the button below. Please note that doing so will share data with third-party providers.