My view of the world, and how the general structure is where I work:
ML is ml. There is a slew of really complex things that aren’t just model related (ml infra is a monster), but model training and inference are the focus.
Backend: building services used by other backend teams or maybe used by the frontend directly.
Data eng: building data pipelines. A lot of overlap with backend some days.
Frontend: you spend most of the day working on web or mobile technology
Others: site reliability, data scientists, infra experts
Common burdens are infrastructure, collaboration across disciplines, etc.
But ML is not backend. It’s one component. It’s very important in most cases, a kitschy bolt on in other cases.
Backend wouldn’t have good models without ML and ML wouldn’t be able to provide models to the world reliably without the other crew members.
The fronted being charts is incorrect unless charts are the offering of the company itself