In every company and even in every DevOps project, processes are often built from scratch – and the smaller the project, the more careless they are. The entire codebase and approaches are collected from articles from the Medium, Stack Overflow responses, and from an employee's previous experience, without any detailed analysis and discussion of the system.
Because of this, most companies spend a significant amount of resources to reinvent a wheel from year to year.
It leads to the fact that any update of the infrastructure becomes a problem. Workers spend most of their time maintaining infrastructure, or not touching it at all. And after a few years, it becomes a terrible legacy code.