In the ML/AI development space, there is what they call MLOps. It has parallels and mirrors DevOps which is widely known and implemented in the software development space. They both have the same purpose of delivering software reliably and sustainably. Something must be said though, that both of these practices require resources and skills that span multiple disciplines and roles.

The tooling and principles between DevOps and MLOps even though they have similarities are different from each other. Hence you cannot assume that DevOps teams can also do MLOps since they have different goals and objectives, but you must ensure that there is an intersection between them.

If you are an organization that deals with both ML and software, you need to be efficient in both DevOps and MLOps, and you need multiple teams to handle them. That is why in some organizations if they start as a software development company, they focus on DevOps and move into MLOps when they integrate AI/ML into their products as you can see with big tech. Or if they are an AI/ML company, they start with MLOps and then move into DevOps when they begin distributing their models into production. In any case, if you are starting either MLOps or DevOps, it is important to master one first before moving to another to ensure focus and alignment throughout the organization.

Reference: https://devops.com/mlops-vs-devops-whats-the-difference/

---