I wonder how ML work is split between ML researchers and software engineers. In particular: - modeling (architecture, loss, hyperparameters) - features (pipelines, transformations) - serving (infra, APIs) Is there ML platform team? Do they work on models at all?
love to know as well