Today, AWS announces two methods to integrate Amazon Aurora PostgreSQL databases with Amazon Bedrock to power generative AI applications. First, Amazon Aurora ML now provides access to foundation models available through Amazon Bedrock directly through SQL. Second, Knowledge Bases for Amazon Bedrock now supports Amazon Aurora as a vector store for Retrieval Augmented Generation (RAG). 

Amazon Aurora ML exposes ML models as SQL functions, allowing you to use standard SQL to pass data to models and return model output as query results. As an example, Aurora ML and Bedrock together can enable real-time summarization of customer support notes in Aurora to accelerate case resolution. Amazon Aurora is also now a vector database option for Knowledge Bases for Amazon Bedrock, letting you securely connect your organization’s private data sources to foundation models for RAG. Through Knowledge Bases, you can also choose to add Amazon Aurora to Agents for Amazon Bedrock to execute multistep actions for your generative AI applications.

The Aurora ML integration with Amazon Bedrock is available in the US East (N. Virginia), US West (Oregon), Asia Pacific (Singapore), Asia Pacific (Tokyo), and Europe (Frankfurt) regions. Knowledge Bases for Amazon Bedrock with Amazon Aurora PostgreSQL is available in the US East (N. Virginia) and US West (Oregon) regions.

To get started with Aurora ML, customers should install the Aurora ML extension and follow these instructions. To get started with Amazon Bedrock, customers should navigate to Amazon Bedrock in the AWS console. To learn more, visit the Amazon Aurora webpage or Amazon Bedrock webpage.