-
Outcome
-
Resolution: Unresolved
-
Undefined
-
None
-
None
-
Market Problem
-
17% To Do, 33% In Progress, 50% Done
Customer need: Customers and prospective customers need an AI platform that provides functionality to address their needs across the AI/ML lifecycle. When evaluating AI platform options, prospective customers evaluate OpenShift AI against competitors that offer solutions that address the complete AI/ML lifecycle.
- Context: slide covering AI/ML lifecycle
- Success Criteria - deliver the following functionality:
- Data Science Pipelines v2 - provide functional parity with all capabilities available in Kubeflow Pipelines v2 in a UI integrated into the RHOAI dashboard.
- OOTB workbench images for VS Code and RStudio
- Plugins for the following IDEs that integrate with RHOAI: VS Code
- Distributed workloads GA
- OOTB model monitoring for performance & ops metrics including GPU utilization
- Model serving enhancements
- Canary rollouts
- A/B testing
- Auto scaling
- Make KServe default model server
- Model registry GA
- Feature store tech preview
- TBD: Model monitoring
- Data drift detection metrics and visualization
- OOD metrics and visualization
- TBD: Hyperparameter tuning