The Watson OpenScale enables responsible, transparent, and explainable AI.
With Watson OpenScale you can:
Evaluate machine learning models for dimensions such as fairness, quality, or drift.
Explore transactions to gain insights about your model.
Watson OpenScale legacy offering plans
Copy link to section
Important:
The legacy offering plan for Watson OpenScale is available only in the Frankfurt region. In the Dallas region, the watsonx.governance plans are available instead.
Watson OpenScale Standard v2 plan
Copy link to section
Watson OpenScale offers a Standard v2 plan that charge users on a per model basis.
There are no restrictions or limitations on payload data, feedback rows, or explanations under the Standard v2 instance.
The regional availability for every service can also be found in the IBM watsonx catalog.
Quota limits
Copy link to section
To avoid performance issues and manage resources efficiently, Watson OpenScale sets the following quota limits:
Asset
Limit
DataMart
100 per instance
Service providers
100 per instance
Integrated systems
100 per instance
Subscriptions
100 per service provider
Monitor instances
100 per subscription
Every asset in Watson OpenScale has a hard limitation of 10000 instances of the asset per service instance.
PostgreSQL databases for Watson OpenScale
Copy link to section
You can use a PostgreSQL database for your Watson OpenScale instance. PostgreSQL is a powerful, open source object-relational database that is highly customizable and compliant with many security standards.
If your model processes personally identifiable information (PII), use a PostgreSQL database for your model. PostgreSQL is compliant with: