Generative AI with DynamoDB zero-ETL to OpenSearch integration and Amazon Bedrock

I will share an experience setting up DynamoDB zero-ETL integration with Amazon OpenSearch to facilitate a natural language query of a product catalog. I will create a pipeline from a DynamoDB to OpenSearch, create Bedrock Connector in OpenSearch, and query Bedrock using OpenSearch as a vector store

In this workshop you will have a hands on experience setting up DynamoDB zero-ETL integration with Amazon OpenSearch Service to faciliate a natural language query of a product catalog. You will create a pipeline from a DynamoDB table to OpenSearch Service, create an Amazon Bedrock Connector in OpenSearch Service, and query Bedrock leveraging OpenSearch Service as a vector store. At the end of this lesson, you should feel confident in your ability to integrate DynamoDB with OpenSearch Service to support context aware reasoning applications.

Pairing Amazon DynamoDB with Amazon OpenSearch Service is a common architecture pattern for applications that need to combine the high scalability and performance of DynamoDB for transactional workloads with the powerful search and analytics capabilities of OpenSearch.

DynamoDB is a NoSQL database designed for high availability, performance, and scalability and focused on key/value operations. OpenSearch Service provides advanced search features such as full-text search, faceted search, and complex querying capabilities. Combined, these two services can satisfy a wide variety of application use cases.

This workshop will allow you to set up one such use case. DynamoDB will be the source of truth for product catalog information and OpenSearch will provide vector search capabilities to enable Amazon Bedrock (a generative AI service) to make product recommendations.

I’ll leverage AWS workshop provided here https://catalog.workshops.aws/dynamodb-labs/en-US/dynamodb-opensearch-zetl