I'm Not Familiar With the Phrase Reasonable Scale.

99% of the campons are not doing much learning at hyper cli. A lot of things can be weged simpler. Not necessarily going ite architecture with practices coming from the biggest players, quite likely. It's not a good idea for most. The average small business that could benefit from machine learning and production doesn't need to do so at five thousand transactions per second. But medium level, even medium level, processing millions of transactions on daily basis, is still not so huge. And youd quite like you wouldn't need to retrain your model every eight hours.