Predictive analytics for DevOps means automating and measuring ‘everything’ are two fundamental DevOps principles for enabling to deliver features and fixes faster. There are good metrics for measuring the speed of value deliveries. Deployment frequency indicates how fast a team can deliver software to production. Lead time metrics provide useful insights on how fast a team can translate requirements into code and deploy them to users. However, looking at the speedometer alone is not enough. It’s easy to deliver something fast but much harder to deliver fast but with quality. Therefore DevOps teams need to build quality in and invest in automating integration, testing and deployment to obtain continuous feedback on functional and technical quality.
So how do you know if you have “quality built in”? In Qentinel Pace, predictive analytics for DevOps uses data from several sources and uses this timeseries data to model the whole software development stack. AI and data is used to find out what are the most influencial predictors of the software quality, what quality prediction is and confidence level of the prediction.