Foundations of Generative AI with Bedrock

This unit introduces the current generative AI landscape, prompt engineering best practices, an overview of the Amazon Bedrock API and multimodal foundational models, and essential security/privacy and project planning concepts. Learners will also install and verify the local development toolchain (Python, Jupyter Notebook, LangChain) and the AWS CLI to prepare for hands-on development in subsequent units.

Learning Objectives

• Analyze the current generative AI landscape by comparing at least three foundational model families (e.g., text, image, multimodal), their typical use cases, strengths, and deployment implications in cloud environments
• Apply prompt engineering techniques to design and iteratively refine prompts for at least two modalities, and assess output quality using criteria such as relevance, correctness, and safety
• Demonstrate the ability to prepare a development environment by installing and configuring Python, Jupyter Notebook, LangChain, and the AWS CLI, and execute a basic Amazon Bedrock API example to obtain and interpret a valid response
• Evaluate security, privacy, and organizational considerations by producing an end-to-end project plan that identifies data handling practices, mitigation strategies, compliance checkpoints, and deployment steps for a generative AI solution using Amazon Bedrock