Today at TrailblazerDX, Salesforce’s developer conference, Salesforce (NYSE: CRM) announced the availability of Einstein 1 Studio, a set of low-code tools that enables Salesforce admins and developers to customize Einstein Copilot — the conversational AI assistant for CRM — and seamlessly embed AI across any app for every customer and employee experience.
Einstein 1 Studio includes Copilot Builder for creating custom AI actions to accomplish specific business tasks, Prompt Builder for building and activating custom prompts in the flow of work, and Model Builder, where users can build or import a variety of AI models. This enables businesses to deliver trusted AI experiences across Salesforce’s Einstein 1 Platform that are tailored to their customers’ needs.
Why it matters: Enterprises face critical challenges in unlocking the power of AI across their business, with 9 in 10 IT professionals saying generative AI has forced them to change the way new technology is implemented and used. They need intuitive user interfaces that make it easy to interact with AI in the flow of work; AI models to fit their use cases; and access to trusted customer and business data to ground the AI models and ensure accurate, relevant outputs.
Salesforce’s Einstein 1 Platform integrates the user interface, a variety of AI models, and data in a single metadata-driven platform. This is what powers Einstein 1 Studio’s tools, facilitating low-code and no-code customization of Einstein Copilot, as well as building and modifying embedded prompts and actions that seamlessly connect to AI models in the flow of work across every Salesforce app. Einstein 1 Studio is deeply integrated with Data Cloud, which safely unlocks and unifies trapped data and grounds AI models with a comprehensive understanding of customers’ data and metadata.
every Salesforce app and workflow. This will help companies boost productivity, improve customer experiences, and increase margins. The tools include:
- Copilot Builder: Create AI actions to accomplish business tasks (beta) — Copilot Builder helps every company configure and customize Einstein Copilot for their business. Salesforce admins and developers can use tools they already have, like Apex, Flow, and MuleSoft APIs, and new generative AI components like prompts, to enable Einstein Copilot to complete tasks in the flow of work. Einstein Copilot can leverage these custom actions to complete tasks across any Salesforce application or external system.
- Prompt Builder: Craft custom and trusted AI prompts with ease (GA) — Prompt Builder empowers admins and developers to create custom, reusable AI prompts without coding, simplifying complex processes and driving business innovation. This not only broadens the use of generative AI beyond conversational interfaces, but also allows customers to design and repurpose prompts for use across other experiences. For example, a custom prompt can be seamlessly embedded in a contact record as a button, enabling an agent in the contact center to get a snapshot of all escalated cases for a customer in one click.
- Model Builder: Choose an LLM or build an AI model based on the job to be done (GA) —Unlike other solutions that limit businesses to a single Large Language Model (LLM), Einstein 1 Studio provides the flexibility to connect to a variety of AI models. Additionally, Model Builder is a no-code, low-code, and pro-code way for companies to build their own predictive AI models, trained on their Data Cloud data. For generative AI, Model Builder allows customers to select from LLMs managed by Salesforce, or bring their own models. Businesses can use predictive and generative AI models and services from Salesforce partners, including Amazon Web Services (AWS) via Amazon Bedrock and Amazon SageMaker, Anthropic, Azure OpenAI, Cohere, Databricks, Google Cloud’s Vertex AI, and OpenAI, and train or fine-tune select models on Data Cloud data without moving or copying data.
Einstein Trust Layer: Deploy AI you can trust, on your terms: Designed for enterprise AI, the Einstein Trust Layer is a collection of features that help companies benefit from generative AI without compromising security or safety standards. New to the Einstein Trust Layer is customer-configured data masking, enabling admins to select the fields they want to mask, providing greater control. Additionally, the audit trail and feedback data collected from AI prompts and responses is now stored in Data Cloud, where it can be easily reported on or used for automated alerts through Flow and other Einstein 1 Platform tools.
Pricing:
- Customers can access Einstein 1 Studio by purchasing Einstein 1 Editions or by adding it on to Enterprise or Unlimited Editions. Detailed pricing information is available here.
Global availability:
- Prompt Builder and Model Builder are now generally available globally. Copilot Builder is available in beta globally.
- Einstein 1 Studio currently supports data residency in the United States and the English language.
- Model Builder support for fine-tuning LLMs on Data Cloud data will be available later this year, starting with Amazon Bedrock, Google Vertex AI, and OpenAI LLMs.