Skip to main content

ML Ops

Recommendation
Updated
Moved
TRIAL
2022-04-29

What is it

MLOps is a set of practices that aims to efficiently and reliably deploy and maintain machine learning models in production. A machine learning workflow consists of data collection, feature engineering, model design and training, endpoint deployment, and monitoring. A complete set of MLOps practices should address all steps of the workflow, as well as simplify cooperation between data scientists, data engineers, ML engineers and software engineers.

Why we use it

To assert that our machine learning models, at all times, are performing as expected. The MLOps practices make sure that all production models are of high quality, assessed continuously, and re-trained and re-deployed as soon as the assessment metrics start to deviate from the accepted range.

When to use it

Whenever a machine learning model is used in a production setting.

How to learn it