QuestionQ106
VerifiedTesting, Validation, and TroubleshootingA university is developing an AI-powered application composed of several sub-applications, including AI assistants, assignment graders, and internal analytics tools. The university is defining and testing multiple prompts across various foundation models (FMs) and wants to:
- Compare variants of each prompt and select the variant that produces outputs best suited for each specific use case.
- Implement version control for the prompts.
- Test prompt variations and collect audit trails covering prompt changes and prompt usage.
- Maintain consistency while still allowing the prompts to be integrated into the main application.
Which combination of solutions will meet these requirements with the LEAST operational overhead?
Choose two
Community Discussion