The application of Machine Learning on DevOps tools (Jira, Git, Jenkins, SonarQube, Puppet, Ansible, etc.), coupled with production data can significantly reduce the effort and cost of the product life cycle.
We identified the potential of applying Machine Learning techniques to massive data sets for pattern detection, identification of inefficiencies, risks and potential failures in key aspects of DevOps. Through this innovation, we intend to go a step further in providing feedback and establishing the basis for a semi-automatic (potentially fully automatic) adaptation of DevOps environments to solve inefficiencies, risks and problems.
The project’s outcome is to design and validate a Machine Learning based system that continuously analyses the data available from the various DevOps process and offers high value-added feedback focusing on: Improvements in QA, Secure Service Deployment, Systems Management in Production, Problem solving and Prevention of Production Failures.
Machine Learning, Cyber-Physical Systems (CPS), DevOps
DevOps, Machine Learning; Process Automation