Posted On: Aug 4, 2023
Today, Amazon SageMaker announces a new direct integration with Salesforce Data Cloud, allowing customers to securely access their data in Salesforce Data Cloud from SageMaker to build, train, and deploy ML models on SageMaker. Customers can then bring the ML models into Salesforce Data Cloud using Salesforce Einstein Studio to power their ML-driven business applications.
With this integration, you can use the SageMaker-Provided Project template for Salesforce Data Cloud integration to setup and standardize the development environment for data professionals in your organization. Next, you can securely connect using OAuth-2.0-based authentication, and directly access data in Salesforce Data Cloud from Amazon SageMaker Data Wrangler, eliminating the needs for ETL pipeline or durable data copies in S3. You can browse Salesforce objects, author SQL queries to retrieve data, join with other data sources such as Amazon S3, and create features using over 300 built-in transforms in SageMaker Data Wrangler visual interface. You can then train your custom model with SageMaker Training, and register it in Model Registry. Finally, you can use the SageMaker-Provided Project template which deploys the model to a SageMaker endpoint, and sets up an Amazon API Gateway. This allows customers to enable inference invocations directly from the Einstein Studio in Salesforce Data Cloud to power business applications.
Salesforce Data Cloud direct integration is supported in all AWS regions where SageMaker is available. To learn more, see the AWS ML blog, Salesforce blog, and AWS technical document for SageMaker Data Wrangler and SageMaker-Provided Project template for Salesforce Data Cloud integration.