Salesforce Investing In AI, Deep Learning

7 Reasons To Convert To A Private Cloud

Salesforce has quietly been amassing talent in the artificial intelligence domain, most recently with the acquisition of MetaMind this week, a company working on deep learning for automated image recognition.

Salesforce CEO Marc Benioff has backed the Palo Alto, California-based startup almost from the beginning, participating in an $8 million venture round in December 2014 along with Kholsa Ventures.

The deal follows a number of other buys by Salesforce of companies including machine learning startup PredictionIO, data science for enterprises company MinHash, and a "smart" iPhone calendar app called Tempo AI that automatically added context such as contacts and documents to calendar items.

Salesforce has also hired away some of LinkedIn's data science talent.

So what's behind Salesforce's interest in this emerging area? Forrester Research principal analyst Mike Gualtieri told InformationWeek in an interview that Salesforce is keeping pace with consumer-focused Internet giants like Google and Facebook with these acquisitions. The MetaMind purchase is about deep learning and visual recognition of objects, a very difficult problem that data and computer scientists have been working on for years.

"The consumer companies are driving the expectations, and I think Salesforce wants to stay ahead of that," Gualtieri said. He believes that Salesforce will continue to target acquisitions in artificial intelligence, machine learning, and deep learning.

"For Salesforce to be a leader they will have to keep with expectations set in the consumer world," he said.

Learn to integrate the cloud into legacy systems and new initiatives. Attend the Cloud Connect Track at Interop Las Vegas, May 2-6. Register now!

Other enterprise companies are also making that a priority.

Gualtieri noted that IBM Watson and HP Enterprise Haven OnDemand offer enterprises deep learning capabilities, and he said HPE's Haven OnDemand is one of the best-kept secrets in the enterprise. For instance, the deep learning capabilities in HPE's Haven enable the system to recognize the voices of particular individuals.

What may be most interesting about the Salesforce MetaMind deal, Gualtieri said, is if the company exposes the deep learning capabilities to partners in the Salesforce AppExchange via APIs. MetaMind CEO and founder Richard Socher wrote in a blog post this week that his company's technologies would be integrated into Salesforce's services.

"With MetaMind and Salesforce coming together, we'll be able to offer customers real AI solutions with breakthrough capabilities that further automate and personalize customer support, marketing automation, and many other business processes," Socher wrote. "We'll extend Salesforce's data science capabilities by embedding deep learning within the Salesforce platform."

Salesforce executives were not immediately available to discuss the company's strategy for AI and deep learning acquisitions. Terms of the MetaMind acquisition were not disclosed. Salesforce did not even issue a press release about the deal.

In the blog post, Socher provided the only real statement addressing what may be Salesforce's strategy.

"Over the past year and a half, we've been on a mission to empower business users with state of the art deep learning technology to simplify, improve, and automate decision making," he wrote. "And now we'll be able to continue our journey at Salesforce on a much larger scale with the resources and ecosystem of one of the world's most innovative and influential enterprise software companies."

On his personal website, Socher identifies himself as chief scientist at Salesforce. Socher delivered a presentation at Salesforce's Dreamforce conference last year about deep learning and what it means for the enterprise.

MetaMind will discontinue its services for unpaid Web users on May 4 and for monthly recurring users on June 4 and delete any data after these closing dates.