Salesforce's Einstein AI is Finally Here

Salesforce Einstein is now generally available to all Salesforce customers. Einstein is an artificial intelligence-based (AI) assistant designed to leverage customer relationship management (CRM) data to help companies discover, predict, recommend, and automate enhanced business processes.

Announced in September of last year, Einstein takes advantage of Salesforce's deep learning, machine learning (ML), predictive analytics, natural language processing, and image processing to serve as a robotic account manager. For example, you can use Salesforce Sales Cloud and Einstein to determine if the person to whom you're pitching a product actually has buying power. Einstein will scan email interactions with your contact to pick up on phrases such as, "Ask my boss" or "Run that up the chain of command" to let you know if you're wasting your time or not. If Einstein sees these phrases, then it will alert you to the possibility that you're not speaking to a decision maker.

Marketers will be able to take advantage of Einstein to automatically create campaign segments based on previous interaction data. For example, Einstein can predict if a contact will open or click through a message. Einstein will then segment your audiences to separate "window shoppers" from buyers. Einstein will then recommend a different kind of content to deliver to window shoppers to prompt a click and a purchase.

These are very general use cases that aim to simplify the standard tasks of any CRM tool. However, Einstein is also capable of processing billions of data points, repetitions, and images to help you improve your workflow. The tool is flexible and intelligent enough to let you program specific automation and prediction outside of the standard CRM use cases. The tool also learns from your use to recommend improved workflows that are specific to your organization.

Can You See What I See?

Einstein isn't only text-based. The tool also leverages image recognition to generate insights from photos. This is especially helpful for research-based activities. Jim Sinai, VP of Marketing for Salesforce Einstein, demoed a scenario in which a solar panel salesperson is set to contact a lead. In the process of researching the lead, the salesperson realizes he or she doesn't know the lead's specific roof type. With Einstein's image recognition enabled, the tool can run a Google Street View search based on the lead's address to pull an image of the roof, recognize the roof type, and notify the salesperson.

These kinds of company-specific processes can be programmed once to run across the entire Salesforce suite for all users, or they can be run automatically whenever a new lead is entered into the system. Einstein Vision application programming interfaces (APIs) give developers the ability to plug image recognition into any Salesforce-connected app to run these kinds of processes for virtually any use case.

Who Gets Einstein?

All Salesforce customers gain access to Einstein regardless of app or price tier. So, if you're only using Salesforce as your helpdesk software, you'll still be able to leverage AI to improve service processes. However, the more data you tie into Salesforce, the more helpful Einstein will be, Sinai said.

Einstein's APIs can be plugged in beyond the Salesforce suite itself to connect to third-party apps and websites. This lets you pull data from your own e-commerce website, or your corporate email account, to help Einstein make more intelligent recommendations.

The CRM AI Wave

Zoho recently added an AI-based virtual assistant to its CRM tool. The new feature, Zoho Intelligent Assistant (or Zia), is an automation engine designed to deliver unprompted, data-based recommendations to sales staff whenever they are using Zoho CRM. Limited to Zoho CRM, Zia is engineered to detect system usage anomalies, suggest optimal workflows and macros, and advise salespeople on when to contact a prospect, according to Zoho Chief Evangelist Raju Vegesna. Zia generates suggestions based on a salesperson's CRM usage patterns, including what's been working well, what hasn't been working well, and what the sales rep can do to improve CRM use.

And, of course, there's IBM Watson, the granddaddy of ML and AI. Watson is primarily a business intelligence (BI) tool but it's also a virtual agent, an e-commerce tool, a marketing solution, and a game show contestant.

Gartner Research predicts that 85 percent of customer interactions will be managed without a human within the next three years, and within the next year, digital assistants will be able to use technology similar to Einstein Vision to recognize customers by the shape of their faces.

This article originally appeared on PCMag.com.