Salesforce Inc. and Amazon Web Services Inc. are launching a new collaboration that will see them more closely link together over a dozen of their services.

The initiative, which expands upon a long-running existing partnership, was detailed today at AWS re:Invent 2023. The main highlight of the announcement is that Salesforce will make many of its core products available for purchase through the AWS Marketplace. In conjunction, it’s rolling out integrations for the Amazon.com Inc. unit’s cloud that will simplify tasks such as customizing large large models. 

“We’re bringing together the No. 1 AI CRM provider and the leading cloud provider to deliver a trusted, open, integrated data and AI platform, and ensuring we meet massive customer demand for our products on the AWS Marketplace,” said Salesforce Chief Executive Officer Marc Benioff. 

Simplified procurement 

Salesforce is making its flagship Sales Cloud, Service Cloud and Data Cloud platforms available via the AWS Marketplace. Also on the list: Tableau, the MuleSoft toolkit for facilitating data transfer between applications, and Heroku, a cloud platform that developers use to build and host software. 

While at it, Salesforce plans to create AWS Marketplace listings for its Industry Clouds. Those are product bundles that combine the company’s core products with features geared toward specific sectors such as manufacturing. Salesforce offers more than a half-dozen such bundles. 

The company says the products’ availability on AWS Marketplace will make them easier to purchase for the thousands of joint customers it shares with the Amazon unit. Moreover, Bloomberg reported, it will be possible to use AWS credits towards Salesforce purchases. That could unlock cost saving opportunities for joint customers. 

At first, Salesforce is making only a limited number of products available via AWS Marketplace and only to customers in the U.S. It plans to add more products and regions next year. According to Bloomberg, Salesforce may explore similar go-to-market partnerships with AWS rivals Microsoft Corp. and Google LLC down the road.

AI takes center stage

Rewriting an artificial intelligence prompt can sometimes significantly improve the quality of the generated output. In September, Salesforce debuted a tool called the Salesforce Prompt Builder that helps developers find the best version of each AI prompt used by their company’s employees. Those optimized versions can be saved as templates and shared with users. 

As part of its expanded partnership with AWS, Salesforce will enable customers to send prompts created with Salesforce Prompt Builder to large language models hosted on Amazon Bedrock. Additionally, customers can customize those language models using the Einstein Trust Layer. The latter tool makes it possible to fine-tune neural networks on datasets stored in Salesforce without giving those neural networks access to sensitive business information. 

“Salesforce and AWS make it easy for developers to securely access and leverage data and generative AI technologies to drive rapid transformation for their organizations and industries,” said AWS CEO Adam Selipsky. “With this expanded partnership, our joint customers gain powerful new ways to innovate, collaborate, and build more customer-focused applications.”

Salesforce will also use AWS services to upgrade the AI features of its Heroku application platform. Developers deploy workloads on Heroku in the form of containers called dynos. To support developers’ machine learning projects, Salesforce will provide the ability to run dynos on AWS instances equipped with AI-optimized chips.

The list of chips that will be accessible to Heroku customers includes the cloud giant’s internally-developed AWS Trainium and AWS Inferentia processors, which are optimized for AI training and inference, respectively. Third-party graphics cards will be available as well. For workloads that prioritize cost-efficiency, Salesforce plans to offer Heroku dynos powered by AWS Graviton central processing units. 

Streamlined data management

Another set of new AWS integrations is set to roll out for Salesforce’s Data Cloud platform. Introduced last year, Data Cloud can store the information a company collects about its customers in one place to ease analysis. Built-in AI features help marketers scan customer records for useful insights. 

The new integrations will enable companies to move records between Data Cloud and AWS services without writing custom ETL, or extract, transform and load, scripts. A set of access controls rolling out in conjunction will help administrators ensure data is used securely. The controls make it possible to regulate who can access records from an organization’s Data Cloud environment that are stored in Amazon S3.

AWS, meanwhile, will adopt Data Cloud internally to manage the customer profiles it uses to inform its user acquisition efforts. The move is part of a broader initiative that will see the cloud giant expand its use of multiple Salesforce products. 

Image: Salesforce