The Reimagined Trailblazer Community is Here! Learn More.

Use the Einstein Platform

Learning Objectives

After completing this unit, you’ll be able to:

  • Explain how the Einstein Platform allows any admin or developer to build their own smart assistant.
  • List each Einstein Platform product.

Build Your Own Smart Assistant

Like you learned in the last unit, we launched Einstein with out-of-the-box features to turn on in our Salesforce clouds. But we all know that Salesforce admins and developers customize their Salesforce instance for the unique ways that their business interacts with its customers. Fortunately, there are tons of ways to customize Salesforce, but what about customizing Salesforce Einstein?

The Einstein Platform allows all admins and developers to build their own AI-powered assistants for a wide array of use cases. You can use point-and-click and programmatic functionality to build applications that predict anything surfaced through Salesforce. 

All of the Einstein Platform products incorporate one or several of the smart assistant components we told you about in unit 1. Read on to dive into the platform products that allow you to build your own smart assistant.

Einstein Bots

Einstein Bots allow you to build a smart assistant into your "customers" favorite channels like chat, messaging or voice.  Einstein Bots use Natural Language Processing (NLP) to provide instant help for customers by answering common questions or gathering the right information to handoff the conversation seamlessly to the right agent for more complex questions or cases. 

Say you’re an ecommerce business, and your service team gets loads of inquiries, through many different channels. With Einstein Bots you can create chatbots for your business to immediately answer specific, routine questions for customers like updating passwords or order status updates. This results in more time for service agents to work on the complex, nuanced cases, and customers save time by getting their answer fast. And if an Einstein Bot doesn’t have the answer at the moment, your customers can be instantly welcomed with a branded greeting in a chat window, directing them to the right agent to help them, fast. Considering customers live in an instant, mobile, web-driven world, they expect one-to-one service—immediately. If they don’t get a swift response, they can think less of a brand. Einstein Bots eliminates that friction.

Corresponding visual on how Einstein Bots works

Einstein Voice


We provide Einstein Voice to selected customers through a pilot program that requires agreement to specific terms and conditions. Einstein Voice is subject to change and isn’t generally available unless or until Salesforce announces its general availability in documentation or in press releases or public statements. We can’t guarantee general availability within any particular time frame or at all. Make your purchase decisions only on the basis of generally available products and features.

Einstein Voice enables all users to talk to Salesforce from any device. Einstein Voice is broken down into two buckets: enabling your organization (Einstein Voice Assistant), and enabling your customers (Einstein Voice Bots), with a smart assistant they can talk to.

Einstein Voice Assistant

Using Einstein Voice Assistant, you can enable anyone in your organization to talk to Salesforce.

Let’s say you have a killer sales organization. Your reps are busy showing off the product, shaking hands, and closing deals, but they’re still not hitting quotas. How can you give them an edge to close deals faster and hit those quotas?

Let’s look into the lens of Sandra—your senior account executive in the Southeast US Region. Every morning when she wakes up, she opens her email to prioritize the accounts, opportunities, and meetings most needing her attention for that day. What if she was told what to prioritize?

The Einstein Voice Assistant can tell Sandra every morning (in place of her lying in bed reading through email) what to prioritize and focus her attention on, so that instead of figuring it out herself, she can use that time to get out of the bed earlier, prepare for those meetings, and be ready for the day.

Fast-forward 3 hours. Her Einstein Voice Assistant tells her about a meeting with a prospective customer, Sharper Fish, LLC. The meeting went great. It seems like she’s about to close the deal, but she just needs to ask one of her product managers a clarification on product functionality. Normally, Sandra has to write or type notes and action items, which not only takes time but can get lost on her desk. With Einstein Voice Assistant, Sandra can talk to Salesforce using her smartphone while she’s driving back from the meeting and log notes. Einstein will understand action items, remind her to do them, and schedule follow-up meetings for her. 

With the help of Einstein Voice Assistant, Sandra was able to close the deal with Sharper Fish, LLC from a quick turnaround on the product functionality question. The next week, Sandra has a meeting with her manager on how the past month has progressed with deals, and what’s left in the pipe. Since she’s so busy selling during the day, she usually spends time during these meetings finding records and navigating dashboards. This is yet another example of time Sandra and her manager can save. Luckily, with the Einstein Voice Assistant, Sandra can drive Salesforce with her voice—creating and navigating through dashboards, and pulling records, all in real time. 

graphic on what Einstein Voice Assistant does

Einstein Voice Bots

With Einstein Voice Bots, your customers can interact with your brand with their voice.

Say you’re an ecommerce business that sells wallets, and you get a ton of inquiries from customers asking about the last status of their orders. Now with Einstein Voice Bots, your company can declaratively build a branded experience to allow customers to speak into their smart speakers  and get updates on their order status, without having to pick up the phone or log into a portal. The voice assistant is integrated with Salesforce, so the customer can ask a question (say about order status), Einstein Voice Bots will search through Salesforce and find the answer, and it will speak the answer back to the customer. This makes it even easier for your customers to get the responses they’re looking for, and it allows your service agents to solve cases faster, leaving the more complicated, nuanced cases for them to focus on. 

graphic on what Einstein Voice Bots does

Einstein Prediction Builder

Einstein Prediction Builder is a simple point-click wizard that allows you to make custom predictions on your non-encrypted Salesforce data, fast. You can create predictions for any part of your business—across sales, service, marketing, commerce, IT, finance, and even HR—with clicks, not code. 

When it comes to understanding how to apply Prediction Builder to your business, ask yourself which objects and fields you want to predict.

  • Do I want to predict the answer to a yes or no question? (Binary Classification)
    • Is this zip code a good opportunity for my business?
    • Will this customer attrit?
    • Does a new employee require a particular type of training?
    • Will a flight arrive on time?
    • Will a customer miss a payment?
  • Do I want to predict an amount? (Regression—in Beta)
    • For what price can we sell this home for?

Say you’re a consumer goods company that sells to retail stores, and you’ve been having a large problem recently with some stores terminating their purchases of your products. Because of that, you want to provide a prediction for your service reps on whether a retail store will attrit. With Prediction Builder, you can choose the object you want to predict on, in this case “retail store” and the field you want to make a prediction for, in this case “attrition?”. Then, you bring in the dataset that includes all of the line items for retail store and whether or not they’ve attrited, along with other characteristics surrounding the customers, and AutoML will do its magic in the backend. 

Finally, Prediction Builder will provide you with a percent likelihood of whether a customer will attrit (for all of the customers who have a blank field for “attrition?”), and the top positive and negative features surrounding the prediction.

Example of a prediction surfaced through Salesforce console

Now that your service reps have this valuable information, they know which customers to engage with, and they have a better idea on how to engage with them to prevent attrition. Ultimately, this will lead to fewer lost opportunities and more revenue for your business.

Einstein Next Best Action

Einstein Next Best Action (NBA) allows you to use rules-based and predictive models to provide anyone in your business with intelligent, contextual recommendations and offers. Actions are delivered at the moment of maximum impact—surfacing insights directly within Salesforce.

Let’s bring back the same scenario where your business was struggling with customer attrition, and you had already built a predictive model for your service reps to see which customers were likely to attrit versus others. Now, what if you could give your service reps the right recommendations to offer the customer to keep them from attriting?

With Next Best Action, you create rules, or propositions, based off of predictions and outcomes, to surface the best recommendation for your service reps to recommend to customers. Say you create a proposition on recommending a 10% discount on two-year contract extensions to customers who have an 80% or higher likelihood of attrition. As your service reps come across predictions of customers who fall in that category, they’ll be recommended right in the Lightning Console to follow up with them and send the 10% discount, and they can take that action instantly, all from Salesforce.

Creating a proposition inside Salesforce console

Example of creating a strategy that takes filters out low likelihoods of attrition

The thing is, there are so many different combinations of follow-up tactics to create for different scenarios for all businesses, so having propositions automatically surfaced in Salesforce, straight to your users helps take the guesswork out of their day.

Example of recommendations in Salesforce console

Einstein Discovery

Like Einstein Prediction Builder, Einstein Discovery also predicts outcomes without requiring your own data scientist.

Let’s go back to the problem of customer attrition in the Prediction Builder example. Let’s say your consumer goods business has some analysts who work with different teams to optimize operations. And your business has troves of data housed in Salesforce with strict data requirements. There are important fields in your Salesforce instance that can help predict customer attrition. Your service reps, who are receiving predictions through the Lightning Service Console from Prediction Builder, start to realize that a few customers are at risk of attrition. Prediction Builder tells them some high-level reasons why, but the service reps want to get to the bottom of the problem. 

With Einstein Discovery, anyone can get the full understanding of relevant patterns on all of the data in your company, whether encrypted or not, to make predictions on customer attrition. You can have full control of the data they’re putting into the predictive model and be able to dig deeper into the predictions and insights. 

For example, Einstein Prediction Builder will show the service reps that a customer is likely to attrit because their last purchase was 3 months ago. Also, their store isn’t located in a region where your products are in high demand. With that insight, how does a service rep respond? Which insight is more important than the other? Einstein Discovery can answer that for you. What if it tells you that consumer demand in a certain retail location is more important? With that insight, your service teams can work with sales, marketing, and product teams to design products that meet the demand for that location, or boost marketing in those areas to increase awareness of the products.

example of predicting what will happen and corresponding graph with Einstein Discovery

Einstein Vision and Language

Building AI-enabled applications can be tough because you have to harness and make sense out of unstructured data. Like we mentioned in the first unit, there are so many types of unstructured data in files like images, text, videos, word documents, and audio files. But all businesses could benefit from predictions on this unstructured data, and this is where Einstein Vision and Language come in. Einstein Vision and Language are a set of APIs and services for Salesforce developers to use to add deep-learning capabilities to any application, ultimately allowing end users to classify images and extract meaning from text.

Einstein Vision consists of Einstein Object Detection and Einstein Image Classification. Together, these APIs harness and make sense out of unstructured data from images to help employees classify them at scale. Let’s say you own a camera company with a variety of products. More often than expected, equipment gets damaged. Reps at your business can take photos of the equipment, and with the help of Einstein Image Classification, they’d be able to understand whether the piece of equipment is damaged, where the damages are, and be given an estimate on how much it will cost to repair. This will take the guesswork out of inspecting all pieces of equipment, and it will save your reps a ton of time.

Einstein Object Detection extracts and contextualizes objects in images. For example, say you’re a company who has loads inventory in warehouses. Your teams can take photos of the inventory, and Einstein Object Detection can identify how many of certain items there are, so your team can accurately plan when to order more, saving on unnecessary spending.

graphic on how to use Einstein Vision

Einstein Language consists of Einstein Sentiment and Einstein Intent. Together, these APIs harness and make sense out of unstructured data from text to help better understand your customers. Let’s say you work for a clothing company, and it has launched a new line of sweatshirts. You want to help your marketers understand how customers feel about your new line of sweatshirts. With Einstein Language, you can build an application that takes in information about the line of sweatshirts, like: what the sweatshirt line is called, what colors are included, what sizes are included, the materials, the locations it has been sold in, etc. Then, Einstein can surface through social media to see whether people post about the product and what is said about the product. Using positive and negative sentiment filters from Einstein Sentiment, your marketers understand who likes or dislikes the sweatshirts, and why they do, so that they can adjust their marketing tactics accordingly. Using Einstein Intent to categorize different text, your marketers can categorize what customers are saying about the product, whether they’re talking about the color, texture, durability, and more. This knowledge inherently your team become better marketers and better sellers.

graphic on how to use Einstein Language

Ready to Get Your Salesforce Einstein On?

Now you’ve seen some of what Salesforce Einstein has to offer. Continue to explore topics of interest to you by checking out the Resources section below.


Rights of ALBERT EINSTEIN are used with permission of The Hebrew University of Jerusalem. Represented exclusively by Greenlight.

Keep learning for
Sign up for an account to continue.
What’s in it for you?
  • Get personalized recommendations for your career goals
  • Practice your skills with hands-on challenges and quizzes
  • Track and share your progress with employers
  • Connect to mentorship and career opportunities