Amazon Bedrock and GenAI Services

This unit explores AWS services for generative AI with a deep focus on Amazon Bedrock, model customization, prompt and flow management, retrieval-augmented generation (RAG), agents, and security/guardrails. It builds on foundational GenAI concepts and prepares learners to design, evaluate, and secure production-ready GenAI components on AWS in the final unit.

Learning Objectives

• Analyze and compare at least three AWS generative AI services (including Amazon Bedrock and Amazon Q) and their roles in an end-to-end GenAI architecture, citing strengths, limitations, and suitable use cases
• Apply Amazon Bedrock features to design and implement prompt management, Bedrock Flows, and agent configurations, producing working templates and flow definitions for a sample use case
• Demonstrate how to build a retrieval-augmented generation pipeline—using knowledge bases, GraphRAG patterns, data automation, and Amazon Q—by ingesting sample data, configuring retrieval, and producing context-aware outputs
• Evaluate model behavior and system readiness by performing model evaluation in Bedrock, implementing guardrails and security controls, and producing a prioritized remediation plan to address performance, bias, and safety concerns