MLOps is like having two partners in one
Consisting of terms “ML + DevOps”, MLOps is a name for a company that takes your data models, sets up a proper process for their deployment by following classic DevOps principles, and develops APIs that makes using (and training) your models in real-life conditions possible.
It all usually starts with a PoC or MVP that should:
- Prove to investors (and other stakeholders) that a particular model isn’t only a great idea, but it also solves real-world problems and
- Basically, MLOps sets a conduit between ML models, stakeholders, users, and the market.