Incident Prevention Through Reliable Changes Deployment

Ensuring the reliability of changes deployment is essential to prevent incidents in businesses that strongly depend on software and services. Incidents should be avoided since they may lead to customer dissatisfaction, financial losses and reputational damage. Currently, the majority of outages are being caused by changes, so we believe there is a need for a higher focus on the risk management pre-change deployment. This paper presents a research plan that proposes a risk management AIOps framework utilising real-world change, CI/CD pipeline and incident data for incident prevention through reliable changes deployment. This research will explore 1) obtaining background information on the current state of practice of service management with a case study on a software-defined business; 2) a risk management AIOps framework that utilises the traces of change, incident and CI/CD pipeline code for predicting the risk of changes deployment; and 3) testing the generalisability of the framework for reducing the risk of change deployment.