This video from AWS re:Invent 2023 explores customizing foundation models for generative AI applications using Amazon Bedrock. Presenters discuss the need for customization, compare fine-tuning and continued pre-training approaches, and demonstrate how to customize models on Bedrock. A case study from the New York Stock Exchange highlights practical applications. The session includes live demos of fine-tuning and continued pre-training using Bedrock's APIs and tools. Viewers will learn strategies for adapting large language models to specific business use cases and domains to improve performance and relevance.