Data Platform Team or ML Platform Team?

For data platform team, I would say that you still need to at least have that first initial team split into two teams responsible for different parts of the product. And when it comes to ML platform teams, the same rules applies or yes, but actually this would not be so soon in the life cycle of the company. Three line teams should eventually also have a machine learning engineer or a data scientist in those teams. So eventually, they will build their own mini platforms inside of teams that's unavoidable. That's also one of the biggest problems that the MLOps platforms try to solve and don't allow teams to diverge too much force.