Definition
The dedicated system and infrastructure responsible for reliably delivering and scaling the real-time prediction or inference outputs of a machine learning model to user applications.
In short: The dedicated system and infrastructure responsible for reliably delivering and scaling the real-time prediction or inference outputs of a machine learning model to user applications.