I use them too, but for data science rather than machine learning. The right abstractions include handling authenticating to the ML services, experiment trackin

I use them too, but for data science rather than machine learning. The right abstractions include handling authenticating to the ML services, experiment tracking, hyper-parameter optimization, distributed compute, access to PII data, etc. If you're dealing with nonsensitive data that fits on a laptop you might not need these.