IBM Watson OpenScale is an open platform that helps remove barriers to enterprise-scale AI. Watson OpenScale enables the enterprise to:
- Measure performance of production AI and its impact on business goals
- Track actionable metrics in a single console
- Explain AI outcomes
- Detect and mitigate harmful bias to improve outcomes
- Track drift
- Accept feedback to compute accuracy measures
- Accelerate the integration of AI into existing business applications.
The goal of this lab is to familiarize the user with the features of Watson OpenScale. Upon completing this lab, you will understand how to:
- Import a machine learning model
- Deploy the model
- Provision Watson OpenScale
- Configure the payload logging database and Machine Learning provider
- Score Data
- Prepare Deployed Model for Monitoring
- Configure Payload Logging
- Configure Quality
- Configure Fairness
- Configure Drift
- Submit Feedback and View Quality Metrics
- Score Data and View Fairness Metrics
- Explain a Transaction.