Salesforce unveils Einstein AI to help close deals

SAN FRANCISCO — In the consumer universe, artificial intelligence is best known as a nice-to-have if stealthy feature that can suggest movies or book rides.

But AI's greater worth could well be in the money-making enterprise arena, where sales, service and marketing initiatives stand to be streamlined by the data-crunching deductive power of machine learning.

That's certainly the bet customer relationship management giant Salesforce is making by unveiling Einstein, the no-brainer name given to a suite of advanced AI capabilities. With Einstein, salespeople can focus on leads that statistically show the most promise of becoming clients, and customer service reps may be better prepared to answer a rainbow of consumer queries.

"The strongest aspect of Einstein is that it is deeply embedded in the platform, it's just working automatically," says Salesforce CEO Marc Benioff, whose company officially rolled out the base product Sunday. More features will be added around the time of Salesforce's annual Dreamforce developers conference in early October.

Google's DeepMind to use AI in diagnosing eye disease

Among the tasks Einstein was designed to help with are sales (predicts leads and prioritizes emails), service (pre-populates case fields and suggests responses), marketing (predicts customer engagement and delivers emails during optimal engagement windows) and commerce (recommends products to shoppers and ranks most relevant site searches).

"Simply put, the computer learns from all your data and uses it to build a model, find patterns and predict the future," says John Ball, general manager for Einstein. "It detects patterns that aren't as easy for humans to detect."

Founded by Benioff in 1999, Salesforce was an early pioneer in cloud-based software at a time with companies such as Oracle were skeptical of the cloud's potential.

Now almost every major tech company has pivoted to cloud-based initiatives, a $300 billion business that includes Amazon's dominant Amazon Web Services, Microsoft's Azure and Google's Cloud Platform. Salesforce, Box and others offer services in more targeted areas.

Oracle's Ellison declares war on Amazon in cloud

Most consumers already interact with AI — an increasingly hot buzzword in the tech world — when they use Apple's Siri or Amazon's Alexa to play music or request an Uber. The same goes for services such as Netflix and Pandora, which use machine learning to predict, with occasional input from users, what a consumer likes.

Machine learning is also integral to the growing self-driving car movement, which uses the technology to help a car's computer brain learn from both its environment as well as its driving errors.

Ball says that Salesforce has been working on its Einstein AI services for two years, acquiring eight machine-learning companies in the process. One of them was MetaMind, an AI startup backed by Benioff that came in-house in April.

Zuckerberg may debut 'Jarvis' AI assistant next month

"Imagine you're working customer service for a company that makes soccer shoes, and you get a call from someone who saw a pair they like on Twitter," says Richard Socher, MetaMind's founder who now works on Einstein. "Well, with just a few descriptive words we can provide a sorted Twitter feed that can help surface what that customer saw."

In a virtual demo of Einstein for USA TODAY, Socher pulls up a screen that mimics what that customer service rep would see. One vertical feed suddenly becomes flooded with recently tweeted images of soccer shoes.

A different demo shows how a sales rep will be prompted by to pursue leads that it ranks as more likely to sign up. Einstein's message: "Recent communication suggests this deal is more likely to be won."

Another screen highlights how Einstein can prompt a salesperson to send an email to a potential client that had mentioned a counteroffer by a competitor in a previous email exchange.

Among the decidedly more consequential examples of Einstein's usefulness is one that involves medical care. Socher explains that radiology companies using Salesforce Einstein can surface for hospitals X-rays showing the most dire conditions based on abnormalities detected by the computer. "There, you're talking about the difference between life of death for someone," he says.

Benioff's new philanthropic mission: the oceans

That's likely a proof point that moves Benioff, who, along with his wife Lynne, has donated some $200 million to the UCSF Benioff's Children's Hospital.

But ultimately, the Salesforce CEO sees AI as presenting both a business opportunity and a societal challenge. Put bluntly, if Einstein can find the best leads and set up an email prompt, what's to stop it from one day closing the deal itself? For free.

"We are designing a system that will help people do their jobs better, but the notion of AI taking people's jobs away is a real issue," says Benioff, crediting peers such as Elon Musk for sounding an AI alarm. "We've all seen these movies. You put one and one together and you get two. We have to strongly watch what we're doing."

Follow USA TODAY tech reporter Marco della Cava @marcodellacava