MLOps
The operational practices for training, testing, deploying, and monitoring machine learning models in production.
Think of it like
DevOps for ML — managing the full lifecycle instead of just shipping code once.
Example
An MLOps team monitors model drift, retrains when accuracy drops, version-controls datasets, and automates canary deployments.
How it actually works
MLOps covers data pipeline management, model versioning, experiment tracking, automated retraining, A/B testing, and monitoring for data drift or distribution shift. Unlike traditional software, ML systems can degrade silently if the world changes but your model doesn't.
For product teams
MLOps is the unglamorous work that keeps models from rotting in production.
For engineers
Build reproducible training pipelines, version everything (code, data, models), and monitor for covariate shift.
Read anything AI without the jargon
Look up any term in plain English, or save terms as you read with the free Chrome extension.
Open DecoderAdd to Chrome