Generative AI platform
Amazon Bedrock Prompt Management
Creates, tests, versions, and manages reusable prompts.
Key points
- Stores prompts so teams can reuse them across applications and workflows.
- Supports prompt versioning, testing, and lifecycle management.
- Helps separate prompt changes from application code changes.
- Can include model parameters and prompt variables depending on the workflow.
- Improves governance by making prompt changes easier to review and track.
When to use it
- Choose Prompt Management when prompts need version control and reuse.
- Use it for shared assistant prompts, summarization prompts, and approved task templates.
- Use it when prompt updates must be tested before deployment.
Exam tips
- Prompt Management governs prompt assets; Model Evaluation compares model outputs.
- Prompt templates are not the same as fine-tuning because they do not train model weights.
- Versioning is the clue when exam questions mention controlled prompt changes.
- Use Guardrails separately when the prompt also needs safety and policy enforcement.