AWS Certified Cloud Practitioner

Explanation

Model Evaluation in Amazon Bedrock Console is the correct answer because:

1. Purpose: Amazon Bedrock's Model Evaluation feature allows users to systematically compare different foundation models (FMs) across various metrics including accuracy, speed, latency, and cost.

2. Key Capabilities:

   • Automated Evaluation: Run batch evaluations on multiple models simultaneously

   • Custom Metrics: Define evaluation criteria specific to your use case (like summarization quality)

   • Performance Comparison: Compare models side-by-side on speed (latency), accuracy, and other relevant metrics

   • Cost Analysis: Evaluate cost-effectiveness of different models

3. Why Other Options Are Incorrect:

   • A) Knowledge Bases: This feature is for connecting foundation models to company data sources through RAG (Retrieval Augmented Generation), not for model comparison.

   • C) Guardrails: This feature is for implementing safety controls and content filtering, not for performance comparison.

   • D) Bedrock Pipelines: This is for creating and managing workflows that chain multiple foundation models or processing steps, not specifically for comparative evaluation.

4. Use Case Fit: For a startup developing a text summarization app, the Model Evaluation feature would allow them to:

   • Test Anthropic Claude, Amazon Titan, AI21 models with the same input data

   • Measure summarization accuracy against ground truth

   • Compare inference speed/latency

   • Make data-driven decisions about which model best fits their requirements

This feature helps organizations make informed decisions about which foundation model to use based on empirical performance data rather than assumptions.