Salesforce Data Cloud and Einstein Model Builder are partnering with Amazon Web Services (AWS) and its Amazon Bedrock service to enable the use of Large Language Models (LLMs) in generative artificial intelligence (AI) applications. This integration was revealed by Salesforce’s Daryl Martis and Darvish Shadravan. Salesforce can bring its own LLMs from AWS accounts to power AI applications, while Amazon Bedrock allows rapid experimentation and evaluation of AI foundation models (FMs) from companies such as AI21 Labs and Anthropic.
The integration also includes a range of tools to promote the responsible use of AI. For instance, the Einstein Trust Layer ensures requests and responses between Salesforce and Amazon Bedrock maintain the privacy and security of data, helping to improve reliability and safety. Furthermore, the BYO LLM experience provided by the Einstein Model Builder allows the registration of custom AI models made in external environments like Amazon Bedrock and Salesforce Data Cloud.
Salesforce’s Einstein Model Builder enables users to intertwine their own Amazon Bedrock models within the Salesforce ecosystem, and develop powerful business applications. The BYO LLM feature utilizes the Einstein Trust Layer and Einstein Model Builder to create trusted AI prompts integrated with Salesforce Flows, Invocable Actions, and Apex, and ensure responsible AI use.
One example of how this partnership works in practice is Anthropic’s Claude model on Amazon Bedrock. This can be used to summarize a list of open service cases and opportunities on an account record page. This tutorial shows how to register the Amazon Bedrock Claude 3 model in Salesforce’s Einstein Model Builder, and integrate the prompt template with a field in the Salesforce Lightning App Builder to build AI-powered Salesforce applications.
Amazon Bedrock service offers a managed, serverless environment, where users don’t have to manage infrastructure. This allows the secure integration and deployment of generative AI capabilities into applications using AWS services that businesses are likely already using.
Overall, this collaboration between AWS and Salesforce provides users with the tools to build bespoke AI-powered applications, using AI models grounded in their own data. The combined resources of AWS and Salesforce are aimed at enhancing users’ generative AI capabilities, while ensuring responsible and secure use of AI. Future posts will provide more insight into how this partnership can be used to enhance business applications.
