RAG and Generative AI on AWS

This capstone unit brings together retrieval-augmented generation (RAG) design, AWS-specific implementation, hands-on labs, and real-world case studies to deliver end-to-end GenAI solutions. Students will build, test, and evaluate production-focused pipelines, consider ethical and security implications, and develop strategies for staying current in the rapidly evolving AI landscape.

Learning Objectives

• Analyze RAG architectures and trade-offs by comparing retrieval strategies, vector stores, and prompt/response workflows to justify design decisions for given application requirements
• Apply AWS GenAI services (e.g., Amazon Bedrock, Amazon Q, data lake integrations) and provided notebooks to design and implement a reproducible RAG pipeline and an agentic LLM assistant
• Demonstrate operational readiness by completing lab exercises that deploy, test, and evaluate model responses, retrieval accuracy, latency, and cost metrics for an end-to-end GenAI solution
• Evaluate ethical, privacy, and security risks and future-facing considerations for deployed GenAI systems and propose concrete mitigation, monitoring, and continuous-learning strategies