Discovering Einstein, Salesforce’s artificial intelligence

Salesforce is undoubtedly the most advanced CRM software vendor when it comes to AI. Salesforce’s artificial intelligence, dubbed Einstein, is already deployed on all the suite’s applications, whether Sales Cloud, Service Cloud or Marketing Cloud.

Einstein’s aim is simple: to save users time, help them make better decisions and make them smarter (“Reveal the Einstein in you”). But what does it look like and how does it work in practice?

If you’re looking to learn more about Einstein and the key use cases for Salesforce’s AI tools, this is the article for you.

Einstein GPT, Salesforce’s generative artificial intelligence

Einstein GPT is Salesforce’s generative artificial intelligence. As its name suggests, it is the fusion of Salesforce’s generative AI and that of Open AI. The two organizations signed a partnership agreement in March 2023.

Einsteint GPT therefore combines Salesforce’s in-house developed AI models with models offered by Open AI.

Einstein GPT is deployed in all Salesforce applications: Sales Cloud, of course, but also Service Cloud, Marketing Cloud, Ecommerce Cloud, Tableau and more. It enables users to formulate requests in natural language directly in the application interface. Use cases are many and varied, and are specific to each application.

On Marketing Cloud, for example, Einstein GPT can be used to generate landing pages, write emails and more. Einstein GPT is designed to support users in the production and automation of certain tasks.

Einstein GPT feeds not only the data used in the models (from Salesforce and Open AI), but also your company’s customer data, provided of course that it is stored in Salesforce.

Einstein GPT is marketed as an add-on module (non-public prices), but is included in the most expensive subscription packages. For example, in Sales Cloud, Einstein GPT is included from the “Unlimited” package at $330 per user per month.

Salesforce Copilot, an intelligence assistant for all Salesforce applications

At its 2023 annual conference, Salesforce announced the launch of Einstein Copilot, an intelligent assistant natively integrated into the user interface of all the suite’s products.

Einstein Copilot goes one step further than Einstein GPT, and is a kind of advanced version of it. It not only allows you to ask questions, but can also perform proactive analysis and suggest actions, making recommendations based on your customer data. Salesforce Copilot is designed to help companies make better decisions and structure their action plans.

Copilot integrates with the Einstein 1 platform, which brings together Salesforce’s AI capabilities and Data Cloud, Salesforce’s data platform. Only Data Cloud users can use Copilot.

Copilot Studio, a tool for deploying your own AI-based applications

Einstein Copilot Studio is the suite of tools that enables companies to build and deploy their own AI-based applications. The Copilot Studio lets you fine-tune Copilot with your own prompts, skills and AI models.

Copilot Studio includes 3 components:

• Prompt Builder. This feature lets you edit prompts for finetuner Copilot. If you use ChatGPT, it’s the equivalent of GPTs. The interface is 100% no code.

• Skills Builder. It’s a low-code tool for creating AI-driven automatic procedures to accomplish specific tasks. Skills Builder is a drag-and-drop interface for connecting data sources, pre-built AI models or custom code. It lets you develop AI skills (workflows, in fact) without using code: lead generation, lead qualification, churn prediction, product recommendation, automation of customer service tasks, creation of personalized marketing campaigns, etc.

• Model Builder. It’s a tool that enables companies to use AI without compromising data privacy. It allows you to create custom AI models to benefit from the power of AI while maintaining control over data and its security. The model builder offers a “Bring Your Own Model” approach, enabling data teams to create and use their AI models directly in Salesforce. Custom models are trained on all data connected to Salesforce.

Einstein Trust Layer, the infrastructure for securing sensitive data

Both Einstein Copilot and Einstein Copilot Studio are based on the Einstein Trust Layer, a secure AI architecture that addresses enterprise data governance issues. Specifically, this feature prevents large language models (LLMs) from storing sensitive customer data. It allows you to take advantage of generative AI without incurring any security risks. Einstein GPT Trust Layer ensures that the data used to feed the models (your customer data) never leaves the Salesforce cloud.

The diagram below shows how the Einstein GPT Trust Layer works:

To help you understand the diagram :

• Dynamic grounding” is a technique that enables a language model to understand the context of a word or phrase by analyzing surrounding texts.
• Data masking” is the process of replacing sensitive data with non-sensitive data, also known as “placeholders”, in order to guarantee data confidentiality.
• Toxicity, in the context of generative AI, refers to the production of offensive or inappropriate content.

Here are the main Einstein use cases for Salesforce’s main products: CRM (Sales Cloud), customer service (Service Cloud) and marketing (Marketing Cloud).

Einstein use case for CRM (Sales Cloud)

Salesforce Sales Cloud Einstein is designed to help sales reps produce more personalized, engaging content.

Here are 3 interesting Einstein features for CRM:

• Sales support. This tool summarizes all stages of the sales cycle in a side panel, from finding customers and preparing meetings to drafting contract clauses. It automatically updates the CRM, so you can be sure that your sales data is always accurate and up-to-date.
• Commercial e-mails. The tool can automatically generate personalized emails for each customer interaction based on your CR data: follow-up emails, reminder emails, sales emails, etc.
• Call summaries. The tool automatically transcribes and summarizes calls, then defines follow-up actions based on the transcript. This can help improve sales productivity by freeing up time that would otherwise be spent manually transcribing calls and defining follow-up actions.

Einstein use cases for marketing (marketing cloud)

Einstein can be used to increase productivity and efficiency in campaign production, audience/segment design and content creation:

• Segment creation. The tool creates intelligent segments using prompts to query Data Cloud, Salesforce’s CDP.
• Content creation. Einstein can be used to generate email content and subject lines, among other things…but also landing pages, forms, SMS content, etc.
• Segment Intelligence. The aim of this feature is to help marketers better understand the performance of marketing campaigns, customer segment by customer segment.

Einstein use case for customer service (Service Cloud)

Salesforce Einstein for Service can be used to improve customer service and perform certain tasks. This can be useful for improving customer satisfaction, reducing costs, improving productivity and making better decisions.

Here are Einstein’s main features for customer service:

• 
  Service responses
  . The tool can automatically generate personalized responses based on real-time data sources, such as your CRM data.
• 
  Job summaries
  . Einstein can produce summaries of customer requests based on available information and relationship history.
• Call summaries. The tool can produce call transcripts and summaries, then identify the next actions to be taken based on the transcript. This can help improve sales productivity by freeing up time that would otherwise be spent manually transcribing calls and defining follow-up actions.
• 
  Knowledge articles
  . The tool can automatically generate and update articles based on the latest real-time data from support interactions, helping you to build up your knowledge base.

AI features for Slack and Tableau

Salesforce has injected AI into all its products, including Slack and Tableau. Slack is a collaborative platform for professionals, Tableau one of the most advanced Business Intelligence tools on the market.

In Slack, artificial intelligence can be used to produce summaries of conversations and assist you in drafting messages. Einstein GPT also lets you bring relevant customer information into Slack from data stored in Salesforce products and the Data Cloud.

Tableau AI, on the other hand, automates many data analysis tasks. Einstein Discovery lets you create predictive models without code. Einstein GPT lets you get answers to specific questions in seconds, via chat. For example: “the percentage of customers who have purchased product X in the last 2 months”.

Salesforce is clearly one step ahead of its competitors, and has succeeded in integrating AI in an impressive way. And this is just the beginning. The article will certainly need several updates!