Blog post on MLflow + Feast + Kubeflow #6009
Description
Is your feature request related to a problem? Please describe.
I'd like a blog post outlining how Feast operates within the AI/ML lifecycle and is complementary to several technologies, particularly MLflow but also the Kubeflow Trainer/Training Operator and Kubeflow Pipelines.
In short, the blog post should outline how Feast supports feature development, iteration, testing, and serving (i.e., productionalizing a model) and MLflow supports experimentation during model development, hyperparameter optimization, and feature selection. MLflow does not support feature serving and is at its core a Metadata tracker and logger. Feast also has a feature registry but this is different than the MLflow registry because Feast's feature registry should be a superset of features in the MLflow registry, this is because once you've chosen your selected model for production, you should only need to serve a subset of those features.
Describe the solution you'd like
A blog post
Describe alternatives you've considered
N/A
Additional context
N/a