Select the right ML App for your use case
Updated over a week ago

Overview

SAVVI ML Apps are designed for a variety of use cases. After you've named your ML App, you'll be prompted to select which type of ML App best suits your use case:

This article provides more information about what each type of ML App is intended to do so you can select the one that matches your use case.


Prediction-Only ML App

Prediction-Only ML Apps allow you to define the influencers that drive the predictions. Unlike the others, Prediction-Only ML Apps do not provide a decision.

A Prediction-Only ML App is a great solution if all your use case needs is a prediction.


Standard ML App

Standard ML Apps allow you to define what the possible decision options will be before the ML App is implemented. You can also define the goal(s) and influencers for the decisions. Because this ML App provides decisions, you can add Guardrails to lock in any business rules that should default certain decisions.

A Standard ML App is a perfect solution if the decision options in your use case rarely change, if ever.


Advanced ML App

Advanced ML Apps allow you to define the attributes across all possible decision options so that the list of decision options can be dynamically retrieved after the ML App is implemented. The Advanced ML Apps take into account the goal(s), influencer(s), and guardrail(s) of your use case.

An Advanced ML App is an ideal solution if the list of possible decision options changes cyclically or unpredictably.

Did this answer your question?