Salesforce AI Announcements from Dreamforce 2023

1. Einstein 1

It all starts here.

Einstein 1 is the foundation to AI in Salesforce.

This will provide a Conversational AI layer within Salesforce for internal users (or embedded in Experience Cloud for external users), enabling users to interact with Salesforce and data housed in it via a conversational interface - Co-pilot.

2. Einstein Co-Pilot Studio

Einstein Copilot Studio is a potentially massive release specifically for Administrators and Developers building within the platform.

This will provide the ability to embed generative AI capabilities into existing flows, automations, and functionality.

Einstein Co-Pilot Studio has 3 primary elements:

Prompt Builder

The ability to create prompt templates, specify how you want those prompts to interact with your Salesforce data, and easily deploy generative AI capabilities for users.

Skills Builder

This is essentially custom permission sets specifically for your generative AI workflows. It allows Salesforce Customers to embed AI with guardrails, specifying how AI should access data and trigger outcomes based on the user type initiating the action.

Model Builder

The ability for companies to control how they build and train their AI models, based on Salesforce’s proprietary large language models or the ability to tap into a preferred partner like Anthropic, Cohere, Databricks, Google Cloud’s Vertex IX, or OpenAI.

3. Innovations to Slack

Slack is starting to see some new feature releases. Slack Lists are a brand new capability (not so much focused on AI) that enable you to track work, triage requests, and manage cross-functional projects in the flow of communication. Slack will undoubtedly play a huge part in Salesforce's AI plans (it is a chat interface, after all) but the first step is embedding the tool further into critical workflows as opposed to it strictly serving as an internal communication tool.

But Slack AI did come with a round of new announcements as well.

Slack AI Capabilities:‍

• Channel recaps instantly deliver highlights on any channel, so users can cut straight to what’s most important.

• Thread summaries enable users to get up to speed on any thread in just one click.

• Search answers allows users to ask a question and search will not only return results with relevant messages, files, and channels, but share an AI-generated summary too.

4. Marketing Cloud & Commerce Cloud Updates

It's likely that we see the most impressive AI features appear in Marketing Cloud before many other products. Salesforce has already announced plans for 50+ Generative AI capabilities on the product roadmap by the end of 2024.

These are primarily around harnessing data in more powerful ways to assist in personalized content creation, more sophisticated re-targeting, predictive analytics and lead segmentation etc.

5. Data Cloud and Tableau Licensing

At the end of the day, Salesforce's success with AI hinges upon adoption. Salesforce now offers 2 Tableau Creator licenses and Data Cloud licenses (up to 10,000 profiles) for free to any Customer already on Sales and Service Cloud (Enterprise or Unlimited Editions only).

6. Data Cloud-Triggered Flow

Data Cloud embedded within the Einstein 1 platform will be able to power automations from real-time data across various systems in your environment. This is part of Salesforce Data Spaces, which is "a logical partition to organize your data for profile unification, insights, and marketing in Data Cloud. [It's the ability to] segregate your data, metadata, and processes into categories, such as brand, region, or department, and then enable users to see and work on data only in the context of their category."