Amazon grows generative AI efforts with Bedrock expansion for AWS

Amazon Web Service (AWS) is doubling down on its initiatives to be the cloud provider of choice for organizations looking to benefit from generative AI.

At the AWS Summit New York event yesterday, the cloud leader outlined its overall strategy for generative AI and announced a series of iterative updates and incremental new services. The latest round of gen AI updates comes as AWS continues to face stiff competition from rivals including Microsoft, with its Azure OpenAI services, as well as the growing set of generative AI services from Google.

The core for Amazon is its AI foundation model service called Bedrock, which was announced back in April, with support for models from AI21, Anthropic and Stability AI, as well the Amazon Titan models. The list of supported models has now been expanded to include Cohere as well the Anthropic Claude 2 and Stability AI SDXL 1.0 models. Beyond expanded model support, Amazon also announced a new Bedrock agent capability to help make it easier for users to build services.

In addition to the Bedrock updates, Amazon announced gen AI capabilities for its Amazon Quicksight business intelligence service, and a preview of a vector engine for the OpenSearch serverless search service. The cloud giant also used the event to announce the general availability of the AWS Entity Resolution service to help improve data management.

“Generative AI has captured our imaginations for its ability to create images and videos and even generate code, and I believe it will transform every application, industry and business,” Swami Sivasubramanian, VP of databases, analytics and ML at AWS, said during his keynote address at AWS Summit New York.

Bedrock gets new agents

An agent in generative AI is a tool that can help to execute multiple tasks on behalf of a developer.

When Bedrock first launched there were no native agent capabilities, but that’s changing with a preview announced at the AWS event. Sivasubramanian explained that Agents for Amazon Bedrock is a new capability for developers to enable generative AI applications to complete tasks in just a few clicks. He noted that agents can be used to help configure foundation models automatically and help to orchestrate tasks without having to write any code.

Sivasubramanian said that the agent securely connects a foundation model to the right data source through a simple API. The agent can also be used to help automatically convert data into machine-readable format.

“Agents in Bedrock can take action by automatically making API calls on your behalf and you do not have to worry about complex systems and hosting them because it is fully managed,” Sivasubramanian said.

More vectors come to AWS

Vector embeddings are an essential element of generative AI, converting content into mathematical representations to enable context and matching.

Vectors must be stored in a database with vector capability. To that end, Sivasubramanian noted that today AWS offers vector database capabilities for its Aurora PostgreSQL relational database. This vector support, announced back on July 13, ad is enabled via the open-source pgvector technology.

Amazon is now extending vector engine support as a preview feature to its OpenSearch serverless search service as well.

“This vector engine offers simple, scalable and high-performing vector storage and search without having to manage any infrastructure,” Sivasubramanian said.

Business intelligence gets smarter with generative AI

Also of note among Amazon’s AI updates is the integration of generative AI into the Amazon Quicksight business intelligence service.

According to Sivasubramanian, business analysts spend a lot of time and effort developing the right data visuals and reports from business data. He said that instead of struggling with complex formulas and commands, with the new generative AI capabilities for Amazon Quicksight, business analysts can use natural language queries to build the reports, visuals and dashboard they need in less time than ever.

So, while there was no major groundbreaking news at AWS Summit New York 2023, the steady drumbeat of iterative innovations that aim to help make AI practically useful in enterprise settings continues at the cloud leader.

“This is just the beginning, we have a lot more coming this year,” Sivasubramanian said.