Dreamforce 2023 was packed full of innovation and inspiration for our developer community, and you can get a great synopsis of all the relevant announcements in the Dreamforce ’23 Retrospective for Developers. But in this blog post, we’ll focus on AI specifically and take a look at the top innovations from Dreamforce that can help you bring AI into your workflow, and even into your apps.

Introducing the Einstein 1 Platform

The Einstein 1 Platform reaffirms our commitment to bringing AI into every aspect of Salesforce. We know that you as a developer depend on the Salesforce Platform and we want you to know that we will continue expanding it, so that you can always have the latest technology at your fingertips. Thanks to some of the innovation around Hyperforce and Data Cloud, we are able to bring generative AI into Salesforce at an incredible speed.

At the core of the Einstein 1 Platform are our CRM solutions, such as Sales Cloud, Service Cloud, and the Salesforce Platform, that Salesforce Developers have been building with for years. Thanks to the metadata framework, we are able to bring together additional capabilities like the lakehouse architecture provided by Data Cloud and the powerful insight and support created by Einstein.

But it doesn’t stop there — we can also use that same metadata to package and distribute your apps through the AppExchange and take your data further with integrations through MuleSoft. Then, through tools like Slack, Tableau, and Heroku you can access and analyze your data from anywhere with direct connections to the Einstein 1 Platform. It is a great new way to see Salesforce become more extensible than ever.

Einstein Copilot and Copilot Studio

The Einstein 1 Platform sets the scene for a glimpse at what the future of Salesforce will look like. Einstein Copilot is a generative AI conversational assistant that utilizes skills and actions to guide users through interacting with Salesforce. Think of it as a brand-new UI for running a business.

This introduces a whole new development paradigm to Salesforce, one where we are creating smaller pieces of functionality that can be orchestrated by Einstein Copilot, and this is our first look at what Assistive AI can do for us.

These skills and actions are brought together in the Einstein Copilot Studio, which allows you to assemble flows, prompts, Apex, and more into collections of functionality!

There are currently three tools within the Einstein Copilot Studio:

  • Prompt Builder allows you to construct prompt templates using merge fields from records and data provided by Flow and Data Cloud
  • Skills Builder allows you to assemble actions, such as Apex invocable methods, flows, and MuleSoft API callouts, and grant them to an agent
  • Model Builder allows you to bring your own AI models to Salesforce

Together, you will be able to build powerful agents in Salesforce that will be able to use your code to answer questions and assist users.

You can get a first-hand look at this in the Dreamforce Main Keynote with Patrick Stokes, our EVP of Product & Industries Marketing.

Demo of Einstein Copilot Studio in the Dreamforce Main Keynote

Einstein for Developers is now in Open Beta

Einstein for Developers is Salesforce’s AI-powered code generation tool designed to enhance developer efficiency. It is now available in Open Beta and developers can access it through a Visual Studio Code extension and Code Builder, Salesforce’s web-based IDE.

Unlike many AI coding tools trained on general languages, Einstein for Developers is specialized in Salesforce languages like Apex. Hosted securely within the Salesforce environment, it leverages Salesforce’s open-source model CodeGen for program synthesis, offering context-aware code suggestions using the org’s Salesforce metadata, and ensuring optimal data security.

Using the Einstein for Developers extension to generate an Apex class

Einstein for Developers was featured throughout Dreamforce ’23, including the breakout session Unlock the Power of Einstein for Developers, which is now streaming on Salesforce+. The session gives you a deep dive into the tool and showcases how you can get started with it today. You can also get an in-depth look at getting started with Einstein for Developers on our developer blog.

Create AI-powered apps with the LLM gateway

The Einstein Trust Layer has been developed by the teams here at Salesforce to create a trusted and secure way for you to interact with large language models (LLM) from Salesforce. It acts as an intermediary gateway between users and LLM providers, masking personally identifiable information (PII) to protect data privacy and running checks for output toxicity while maintaining an audit trail. The layer abstracts differences between model providers and ensures that no user data is persisted or used for further training by model providers. The Einstein Trust Layer allows you to send requests to LLMs in a trusted way, addressing the concerns mentioned above.

Building LLM apps with the Einstein Trust Layer

Learn more about how the trust layer works in the blog post Building AI-Powered Apps with LLMs and Einstein.

Prompt Builder

One of the simplest ways that you can utilize the trust layer is through the Prompt Builder, part of the Copilot Studio. It is a tool that allows users to create, test, and manage prompt templates while integrating them with Salesforce data. Prompts make it much easier to communicate with an LLM and provide strict instructions on what type of content should be generated.

Prompt Builder helps you embed generative AI capabilities into different Salesforce experiences, such as email and record creation, with many more on the way.

Using the Prompt Builder to generate an email with Salesforce data

Sneak Peek: Bring generative AI anywhere with Apex

Salesforce Developers will be able to build their own generative AI applications with the Einstein 1 Platform. You will soon be able to use the Einstein LLM Generations API to connect to your choice of LLM from Salesforce-approved vendors and integrate with the Einstein Trust Layer out of the box. This functionality is currently in a closed pilot with more details to be announced soon!

You can get a glimpse of this functionality in the Dreamforce Developer Keynote: The Future of Development, where Product Manager Avanthika Ramesh walks us through the foundations of the technology.

The functionality will be exposed through the Einstein Connect API Adapter, and you will be able to generate simple and complex generations that can be accessed anywhere in Salesforce!

Summary

As you can see, Dreamforce 2023 was full of AI innovations for developers. The event introduced the Einstein 1 Platform, emphasizing Salesforce’s dedication to integrating AI into the Salesforce Platform. The Einstein Copilot acts as a generative AI conversational assistant, bringing a new development approach for Salesforce, and the tool is exposed through the Einstein Copilot Studio. Not only this, but our commitment to making it easier for you to build your own AI-powered applications through the Prompt Builder and Einstein LLM Generations API will give you the opportunity to create apps like never before.

We can’t forget Einstein for Developers, which several thousands of you are already using to help you get started with boilerplate code right inside of your IDE.

Be sure to check out all of the content from Dreamforce ’23, and we can’t wait to see what you build with AI!

Resources

About the author

Stephan Chandler-Garcia is the Director of Strategic Content at Salesforce. He has been in the Salesforce ecosystem for over 10 years as a customer, partner, and ISV. You can find Stephan in person at a Trailblazer Community Group or at one of our conferences around the world. Alternatively, follow him on X (Twitter) or GitHub.

Get the latest Salesforce Developer blog posts and podcast episodes via Slack or RSS.

Add to Slack Subscribe to RSS