What is Watson OpenScale?
IBM Watson® OpenScale™ tracks and measures outcomes from your AI models, and helps ensure they remain fair, explainable and compliant — for models built and running anywhere. Watson OpenScale also detects and helps correct the drift in accuracy when an AI model is in production. Access Watson OpenScale in IBM Cloud Pak for Data™ and on IBM Cloud™.
Featured resources
-
Go deeper on AI drift
- Read the documentation
-
Get an IBM data scientist’s POV on Watson OpenScale
- Read the blog (Link resides outside IBM)
-
What is AI model drift and how is it risky?
- See the infographic (PDF, 6.9 MB)
-
Take a tour of drift — and how to mitigate it
- Register to watch the presentation
Operationalize AI
Watch webinar: Faster AI value
IBM experts explore the five best practices that can speed AI value in your organization. One of them is making AI trusted and transparent.
How to keep AI models accurate
See what's involved in monitoring a model for drift and determining the causes of drift.
Be confident in your AI
Learn how Watson OpenScale solves AI's black box problem, helping you make outcomes fair, explainable and compliant.
Take an interactive tour
Explore Watson OpenScale in a self-guided, hands-on demo.
Product images
Detecting drift
Watson OpenScale tracks drift in model accuracy and sends an alert when accuracy degrades past a threshold.
Debiasing AI models
OpenScale generated a bias alert because significantly fewer females than males were approved. OpenScale also automatically generates a corrected, debiased model that can be evaluated and deployed.
Explaining AI outcomes
OpenScale explains the reasons behind each model prediction: this candidate was a risk due to age and no checking account.
