Model A/B testing for recurring assignmentsParzivalPZ-001
Compare two or more AI models on the same recurring assignment. Each model reviews the others against a shared rubric, token use, and user feedback so Parzival can choose the best model for that responsibility going forward.
- Products
- Parzival
- Status
- In development
- Release phase
- Private alpha
- Platforms
- Linux
- Added
- 2026-07-17
- Modified
- 2026-07-17