MLOps is just a buzzword, but maintaining ML projects in large organizations has several major pains, e.g.: http://papers.nips.cc/paper/5656-hidden-technical-d

MLOps is just a buzzword, but maintaining ML projects in large organizations has several major pains, e.g.: http://papers.nips.cc/paper/5656-hidden-technical-debt-in-machine-learning-systems.pdf http://papers.nips.cc/paper/5656-hidden-technical-debt-in-ma... I use MLFlow at my company for solving two specific problems: tracking experiments and model performance. But if you're small enough, you probably don't need MLOps, Devops, etc...