Amazon Bedrock allows you to use of generative AI models in your workflows and applications in a variety of fully-managed and partially-managed ways. Knowledge Bases are a key feature of Amazon Bedrock, enabling you to create fully automated Retrieval Augmented Generative (RAG) models solutions using your own data. Knowledge Bases support using your own custom vector store, which can have many benefits including, cost, privacy, and performance optimizations.

Learning how to configure a vector store for use with Amazon Bedrock will benefit anyone looking to implement a RAG solution using their own data in the public AWS cloud.

In this hands-on lab, you will configure an Amazon RDS PostgreSQL cluster as a vector store for use with an Amazon Bedrock Knowledge Base, and you will create an Amazon Bedrock Knowledge Base that uses it.

Please note: This lab creates an Amazon RDS PostgreSQL cluster which can take up to ten minutes to provision. Please ensure you have enough time available before starting the lab.

Learning objectives

Upon completion of this beginner-level lab, you will be able to:

• Create vector store database objects in an Aurora PostgreSQL cluster
• Update a secret in AWS Secrets Manager to use credentials that you created
• Create an Amazon Bedrock Knowledge Base that uses the vector store you configured

Intended audience

• Anyone interested in using generative AI in the AWS cloud
• Cloud Architects
• Data Engineers
• DevOps Engineers
• Machine Learning Engineers

Prerequisites

Familiarity with the following will be beneficial but is not required:

• Amazon Bedrock
• Aurora PostgreSQL
• AWS Secrets Manager

The following content can be used to fulfill the prerequisites:

• Amazon Bedrock Orchestration using Knowledge Bases and Agents
• Introduction to Amazon RDS & DynamoDB
• Using AWS Secrets Manager to Manage, Rotate, and Retrieve Secrets