Answer: Track the model changes by using Amazon SageMaker Model Cards.

## Detailed Explanation **Correct Answer: C (Amazon SageMaker Model Cards)** Amazon SageMaker Model Cards is the optimal AWS solution for this scenario because it provides a standardized, managed framework specifically designed for documenting and tracking AI/ML models throughout their lifecycle. **Why Option C is Correct:** 1. **Purpose-Built for Model Documentation**: SageMaker Model Cards are AWS's dedicated solution for creating standardized documentation about ML models, including intended use cases, training details, evaluation metrics, and performance characteristics. 2. **Version Tracking Capabilities**: Model Cards inherently support versioning, allowing teams to track changes across different model iterations, which is crucial when collaborating with multiple research institutes. 3. **Standardized Format**: The solution provides a consistent template that ensures all collaborators document models in the same structured way, facilitating clear communication and reducing misunderstandings between different organizations. 4. **Development Process Recording**: Model Cards capture the entire model development journey, including data sources, training parameters, evaluation results, and deployment considerations. 5. **AWS Integration**: As a native AWS service, it integrates seamlessly with other SageMaker components and AWS services, providing a unified platform for model development and documentation. **Why Other Options Are Less Suitable:** - **A (Git)**: While Git is excellent for version control of code, it lacks the specialized structure and features needed for comprehensive ML model documentation. Git doesn't provide standardized templates for model metadata, performance metrics, or compliance documentation that are essential for cross-organizational collaboration. - **B (Amazon Fraud Detector)**: This is a specialized service for fraud detection use cases, not a general-purpose model documentation solution. It doesn't provide standardized documentation capabilities for arbitrary AI models being developed collaboratively. - **D (Amazon Comprehend)**: This is a natural language processing service for text analysis, not a model documentation or version tracking tool. It's designed for extracting insights from text, not for documenting ML model development processes. **Key Considerations for the Scenario:** The requirement involves collaboration with multiple research institutes, which necessitates: 1. **Standardization** - All parties must use the same documentation format 2. **Transparency** - Clear visibility into model development decisions 3. **Auditability** - Traceable record of model changes and rationale 4. **Centralized Management** - Single source of truth for model documentation SageMaker Model Cards addresses all these requirements effectively, making it the most appropriate AWS solution for this collaborative model development scenario.