Salesforce, led by CEO Marc Benioff, has taken a significant stride in the AI landscape with the introduction of Einstein Studio. This groundbreaking platform offers a “bring-your-own-model” experience, enabling enterprises to connect and train their proprietary AI models using Salesforce’s data.
The primary objective of Einstein Studio is to streamline the AI project lifecycle, facilitating data science and engineering teams in managing and deploying models more efficiently, quickly, and cost-effectively. Once trained, these models can be seamlessly integrated into any Salesforce application, including sales, service, marketing, commerce, and IT functionalities.
Rahul Auradkar, EVP, and GM for unified data services and Einstein at Salesforce, emphasized the significance of Einstein Studio, stating that it provides a faster and easier way for customers to create and implement custom AI models. By utilizing their own exclusive data, Salesforce customers can leverage predictive and generative AI across all aspects of their organization.
The offering underwent testing with multiple enterprises as part of a pilot program and is now accessible to all users of Salesforce’s Data Cloud.
How Will Einstein Studio Benefit Salesforce Customers?
In today’s competitive landscape, businesses are racing to deploy AI models targeting various critical use cases, such as predicting future demand or delivering personalized recommendations. However, the process of building and deploying enterprise-ready AI across applications and workflows is highly resource-intensive and time-consuming. Data extraction, transformation, and loading (ETL) are just the initial stages, followed by model training and implementation, which require careful monitoring throughout the project lifecycle. This complexity often hinders teams from deploying projects promptly.
According to a KPMG survey, around 60% of U.S. executives anticipate implementing AI solutions within the next year or two.
Einstein Studio addresses this challenge by expediting the AI deployment process. It enables users to connect custom AI models, built with external services like Google Vertex AI and Google Vertex AI, and train them using data hosted within Salesforce Data Cloud, tailored to address specific business needs.
Salesforce Data Cloud aggregates data from various sources to create unified customer profiles that adapt to each customer’s real-time activity. Einstein Studio’s pre-built, zero ETL integration harnesses this data directly for model training, requiring users to simply “point and click” on relevant data assets within the platform.
Additionally, the BYOM (bring-your-own-model) solution provides a control panel, empowering data scientists and engineers to govern how their data is exposed to AI platforms for training.
Sanjna Parulekar, VP of product marketing at Salesforce, highlighted the benefits of Einstein Studio in streamlining the entire AI project lifecycle, from data acquisition and preparation with Data Cloud to modeling, deployment, and insights consumption. This approach allows organizations to integrate business, IT, data professionals, and end-users fully, maximizing their investment in the latest AI platforms.
Deployment Within and Beyond Salesforce
Once a model is trained using Data Cloud, it can be integrated into various Salesforce experiences, including Data Cloud, Flow, and Apex, to power company applications. For instance, a propensity-to-buy model built using AWS SageMaker and registered in Einstein Studio could be utilized in a Flow automation to decide whether to send a product discount email to a customer.
Furthermore, customers and independent software vendors can also use these trained models in external applications. Retailers, for example, can employ the models to provide personalized product recommendations, individualized pricing based on customer needs, or segment customers according to demographics and purchase history.
David Geisinger, global alliance lead at Deloitte Digital, expressed excitement about Salesforce’s Data Cloud and Einstein Studio, enabling customers to bring their own models and utilize AI and customer data creatively. Deloitte has developed a series of “bring-your-own-models” that clients can leverage within the Salesforce ecosystem.
As of now, Einstein Studio allows users to build custom models from scratch or connect with AWS SageMaker and Google Vertex AI. Salesforce plans to add more services in the future. The offering is automatically enabled for all Salesforce Data Cloud users from its launch date.