A Task
To be solved in 48 hours. Task: We have a recommendation engine dataset stored currently in a csv but will later be transfered to a feature store. We want to build the following pipelines on kubeflow for it.
a. Feature engineering pipeline.
b. Feature store with feast
c. Data versioning pipeline.
d. Model training pipeline.
e. Hyperparameter tunning with katib
f. Model deployment with seldon core.
g. Model monitoring with prometheus and grafana
Implement a Continous integration such that any of the stages can be automatically updated and changes propragated to the pipeline.
This is a basic project and should be completed in at most 48 hours, please do not spend more than 48 hours on this project and we want to assume that the bulk of the work is in setting up your local kubeflow.
When you done, write your approaches and consideration in the Readme.