Most of the data created in the process of DevOps is related directly to the application deployment. Application monitoring replenishes server logs, generates error messages, and transaction tracing. The only reasonable way to analyze this data and make the right conclusions in real time is to use machine learning (ML).