M+E Daily

Salesforce continued its artificial intelligence (AI) and machine learning technology acquisition binge by signing a definitive agreement to acquire Twin Prime, a Redwood City, Calif.-based technology startup that specializes in mobile performance and app acceleration.

Salesforce didn’t say how much it’s paying for Twin Prime and declined to comment on the deal Dec. 14, only referring MESA to a Dec. 13 blog post by Twin Prime CEO and founder Kartik Chandrayana.

In the post, Chandrayana said his company’s team had expertise in mobile, networking and data science, and they developed “a solution that leverages machine learning techniques to analyze real-time network data, identify and prioritize performance gaps in mobile apps, and then optimize mobile performance so companies can deliver content faster to any location, device and network.”

By collaborating with customers and partners, the company was “able to help solve mobile performance problems across almost every network and geography in the world.” he said.

Chandrayana went on to say that, “with Salesforce’s global footprint and scale, we see this as a great opportunity to take our product to the next level and boost the user experience for every Salesforce customer in the world.” The Twin Prime team will “help optimize Salesforce’s global connectivity and network performance through machine learning for all Salesforce clouds and services,” he said. He didn’t project when the purchase is expected to close.

Salesforce has made several acquisitions in the machine learning and AI spaces prior to Twin Prime, including MetaMind, PredictionIO, Implisit, Demandware, Quip, BeyondCore and Gravitytank. Salesforce executives have said they are using technologies from those purchases as part of its new Einstein AI platform.

Salesforce CTO Parker Harris, meanwhile, provided a launch update for its Einstein AI platform Dec. 13, while speaking at a Pacific Crest Securities investor meeting in San Francisco. Sales Cloud Einstein will go into general availability this spring, he said, noting “we have pilot customers now” who are using that.

No specific release dates have been announced yet for the other planned implementations of Einstein, but “we’re working on all of that now,” he said, pointing out that Salesforce makes three big product releases each year: one each in spring, summer and winter, in around February, June and about the same time as its annual Dreamforce conference in October. “With each of those you’ll see progressively more progress on the Einstein platform and more announcements,” he said.

Salesforce will monetize Einstein and the feedback on it from customers so far “has been great,” he said, adding they’re “willing to pay for it” and see the added value of the AI platform
The company is “not doing anything today on voice” as far as any launched products and services, he also said, adding: “I think Google, Microsoft and Amazon all have a lot of technology around voice.” But he said Salesforce does have a research group via its purchase of the Palo Alto, Calif.-based AI company MetaMind early this year and “we’re looking at research around voice – specific to” customer relationship management (CRM).

The main question that will be asked by Salesforce and its research team, he said, is “can we do things that might be better for CRM than maybe a generic voice platform?”