Salesforce's MetaMind Buy Fuels CRM's AI Arms Race


ManageEngine OpManager, a powerful NMS for monitoring your network, physical & virtual (VMware/ HyperV) servers & other IT devices. Deploy and start monitoring in less than an hour. Trusted by over a million admins world-wide. Try it for free.

MetaMind, a startup focused on deep learning, this week announced that it has been acquired by Salesforce.

The company's expertise is in natural language processing -- letting computers analyze relationships between words.

"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," MetaMind CEO Richard Socher said.

The acquisition will see MetaMind, whose backers included Salesforce CEO Marc Benioff, embed deep learning within the Salesforce platform.

Socher has been named chief scientist at Salesforce.

MetaMind's system provides a dynamic memory network that offers a way for machines to answer questions about what's said in a piece of text.

The system also can judge the general feeling the words express, identify parts of speech, determine what a pronoun is referring to and translate between languages, according to the company. It performs those tasks using a question-and-answer approach.

MetaMind built a breakthrough AI platform aimed at letting everyone solve important problems, and its core NLP outperforms benchmarks on several tasks, according to Socher.

Its technology matched Google's 2014 entry for the ImageNet large-scale visual recognition challenge using significantly less compute power, he said.

On a sufficiently large data set, MetaMind's platform will tune the user's classifier for very high accuracy, Socher added. The company provided a simple API so developers could embed their classifiers into their apps with a few lines of code.

Many developers using the company's platform have created public classifiers, such as Sentiment (tweets), that other developers can start using at once.

Access for MetaMind's unpaid subscribers was cut off when the deal was announced; paying subscribers will have access discontinued June 4. In both cases, MetaMind will delete subscribers' data after terminating their access.

"All the leaders in CRM are looking to advance their AI and embedded analytics capabilities as differentiators in the last mile of CRM -- greater productivity for users," remarked Rebecca Wettemann, a VP of research at Nucleus Research.

"The arms race in AI is far from over," she told CRM Buyer, "and we shouldn't forget that the grandfather of AI -- IBM with Watson -- is upping its ante in the CRM battle as well."

Meanwhile, Microsoft Dynamics CRM got AI-powered predictive marketing capabilities in March through integration with EverString's platform.

The integration lets marketers analyze and validate their entire addressable market of prospects, then proactively select and target the optimal audience based on data from their Microsoft Dynamics CRM Online package and more than 20,000 external signals.

"The dark-cockpit principles of modern software UI design bring intelligence of the software to bear, which requires machine learning and AI," Wettemann said.

"What MetaMind built makes it easy to address the service use case, but there's much more," noted Ray Wang, principal analyst at Constellation Research.

"The cognitive field is just starting, and companies are looking to take data, context and engagement to the next level by identifying patterns that help make better recommendations, suggest smarter actions, and identify issues before they occur," he told CRM Buyer.

On the other hand, the purchase "has the earmark of an acquisition done to bring promising ideas in-house without necessarily having a clear path to market," suggested Denis Pombriant, principal at Beagle Research.

"As one of the larger companies on the planet, Salesforce is moving into an area where it's going to need to sponsor primary R&D to keep its growth humming." he told CRM Buyer, "and this could be seen as a first step."

Richard Adhikari has written about high-tech for leading industry publications since the 1990s and wonders where it's all leading to. Will implanted RFID chips in humans be the Mark of the Beast? Will nanotech solve our coming food crisis? Does Sturgeon's Law still hold true? You can connect with Richard on Google+.

0 Comment

Leave a Reply

Captcha image


  • 5300c769af79e

    Google Nexus Smartwatches On The Way?

    Read Google is reportedly preparing two new Android Wear smartwatches, possibly carrying the Nexus brand, for a debut later this year.Following a path similar to that of its Nexus smartphones and tablets, two new wearables are being designed by Google but will probably be manufactured by an OEM partner, according to reports.
  • 5300c769af79e

    IBM Watson Health Tackles Diabetes

    To address the challenge, IBM Watson Health and the ADA are collaborating to apply Watson cognitive computing to the organization's massive library of information and data.It was one of several announcements from IBM aimed at improving research and care for diabetes patients.
  • 5300c769af79e

    The Cost of an Unintentional Insider Threat

    Download A high rate of Unintentional Insider Threats (UIT) sheds light on the demands of today's workplace culture in the United States.Work cultures are unlikely to change anytime soon, but there are tools and programs that can help any work environment with a UIT problem.
  • 5300c769af79e

    #TBT: Bugless Beast, the One and Only

    If I’m not mistaken, Peter Alfonso, the man behind Bugless Beast, hasn’t been into Android development for a bit of time, but at one point a few years ago, everyone knew his name, or at the very least, knew of his work.If you don’t know, Bugless Beast was a custom ROM, downloadable on an assortment of different Android devices and I would safely argue that it was one of the most popular things we would post about back in the day.