# MLOps

> MLOps is the set of practices for deploying, monitoring, and maintaining machine-learning systems in production reliably, the discipline that keeps AI working after launch.

_Source: https://plenaura.com/glossary/mlops · Last updated: 2026-06-03 · Plenaura_

## Key points

- Covers deployment, monitoring, retraining, and performance tracking.
- Prevents the silent model decay that degrades accuracy over time.
- The difference between a model in a notebook and a system that serves users.

## FAQ

### Why does AI need ongoing operations?

Because real-world data shifts, model accuracy drifts without monitoring and occasional retraining. MLOps is what keeps a deployed system trustworthy instead of quietly degrading.