AIOps for CI with Kubeflow Pipelines
summary
An open source AIOps system called AI4CI automates Jupyter notebooks with Kubeflow Pipelines to collect CI data from Prow and TestGrid, calculate concrete KPIs, cluster build logs and run failure type prediction models. The pipeline ingests job metadata and log events, transforms features, trains models, and pushes results to interactive dashboards so teams can see which failures cluster together, which predictors matter, and where to look first to root cause CI breaks.
AI-written summary · about
related talks
44mWhat’s Breaking My Build? AIOps Tooling to Aid Your CI/CD Workflows
34mOptimize Long Test Runtimes Using Open AIOps
05mCI/CD Pipelines and Machine Learning
27mKubeflow for enabling AI Powered Drug Discovery and development...
36mMachine learning pipeline Kubeflow, let's start
42mAI – Addressing the Challenges of Today’s Data Scientists
21mAI-driven optimization of CI/CD pipelines
54mScaling Machine Learning Pipelines in Cloud