Einstein Bot and Einstein Voice are here, and they represent growth milestones for the 3-year-old suite of Salesforce AI features named after the genius physicist.
Einstein -- the Salesforce AI, not the physicist -- was born a little over three years ago in the basement of a furniture store on University Avenue in Palo Alto, Calif.
It was the brainchild of a small team of Salesforce AI enthusiasts who, like many tech entrepreneurs before them, felt that a "back-to-the-garage" approach would fuel innovation. It was there that they created an AI to rival IBM's mammoth Watson.
Now 3 years old, Einstein has proven himself to be as much a prodigy as his namesake.
Growing up fast
Einstein progressed with phenomenal speed after its introduction to the world in 2016. Serving a population of more than 3 million users, its daily number of operations is already in excess of 1 billion.
Some notable features include Einstein Sentiment, which can identify emotion in email and text; Einstein Intent, which can sort through customer inquiries and generate relevant responses to specific problems; and Einstein Object Detection, which can pick out unique objects in images. That's over and above Einstein Analytics, the core functionality it put forth to initially challenge Watson.
A desire to please
The 3-year-old Einstein has a desire to please. The first Salesforce AI tool, originally dubbed Optimus Prime, was a "What's next?" tool for organizing sales work. The Einstein Intelligence module was a people-pleasing timesaver that sorts a salesperson's leads by "most likely to close," along with the ability to sort email lists by those recipients most likely to open an email.
The buildup of features and apps since has followed this pattern -- to not only generate insights and predictions, bolstering efficiency holistically across CRM systems, but to provide personal labor-saving functionality to individuals.
Learning to talk
Einstein likes to talk. This feature of AI is now widely used, thanks to Siri and Alexa, but Einstein, per Salesforce's promises, is talking to users, not at them. Its core natural language processing capability springs from a powerful general model that enables an array of NLP functions -- including Q&A, inference, goal-oriented dialogue and summarization -- available to developers for faster and more efficient exploitation.
More impressive still is Einstein Voice, now in beta, which empowers users to talk to Einstein apps on their mobile devices, with Einstein parsing verbal instructions to retrieve and process relevant records and tasks.
Plays well with others
Salesforce partnered with IBM to bring Watson's insights functionality into Quip and Einstein. IBM is now a preferred cloud services provider for Salesforce, and Salesforce is the preferred customer engagement platform for IBM.
Einstein demonstrates considerable talent at an early age. This talent springs from Salesforce's longtime commitment to codelessness, which led Einstein's creators to confront the age-old conundrum of data science: How does the enterprise get past the problem of the front-end labor required to achieve AI's rewards?
Making AI app creation codeless was only the first step. Industry after industry stalls in implementing AI over the time and effort required to build and implement effective models upon which to build the apps they need.
"There aren't enough data scientists in the world to build all the predictive models we need," said John Ball, Einstein's general manager.
So, Salesforce automated the model building process -- no data scientist required, no coding needed. This puts Einstein at the head of the AI class.
New features to come
Like most 3-year-olds, Einstein shows promise. Each of the features and traits detailed above is impressive in itself, but the exciting part is how they'll be used in combination.
Einstein Bot is codeless, as you'd expect, and easy to configure and deploy, while being flexible in behavior; it can interact behind the scenes with your other Salesforce apps and data.
Over the coming year, Salesforce users will not only move forward with Einstein Bots in sales and service applications, but will also exploit them alongside these other new features -- Voice, Intent, Sentiment and Next Best Action, which uses predictive analytics as the core of its recommendation engine.
The end result will be apps -- many of them bots -- of unprecedented AI utility. The gates of AI are daunting and often frustrating to the enterprise. Salesforce is betting everything that it can change the game.
Put another way -- Einstein is going to be one impressive 4-year-old.