UPDATED 15:00 EDT / FEBRUARY 06 2018

NEWS

The hardware-software-data trifecta in Cisco’s enterprise AI

Artificial intelligence assistants like Amazon Echo combine a hardware device with software models trained on big data. But not many companies have the necessary chops in all key areas to build a highly intelligent device. Could Cisco Systems Inc.’s position as a legacy hardware provider transitioning to a software platform plant the company in the AI-assistant sizzle spot?

“We went from hardware to software; now we’re going from software to being a platform company in many ways,” said Jose de Castro (pictured), chief technology officer of cognitive collaboration and Spark platform at Cisco Systems.

The company is pulling together different elements of its portfolio to offer a range of applications and services for collaboration within enterprises. “We think of ourselves as building the operating system for teams,” de Castro said, who spoke with John Furrier (@furrier) and Stu Miniman (@stu), co-hosts of theCUBE, SiliconANGLE Media’s mobile livestreaming studio, during the Cisco Live event in Barcelona, Spain. (* Disclosure below.)

This week, theCUBE spotlights Jose de Castro as our Guest of the Week.

Spark AI soup

At the center of the platform taking shape is the company’s collaboration suite, Cisco Spark (not to be confused with the open-source Apache Spark platform for managing large data sets). Additional pieces include Cisco’s WebEx video and web conferencing solution and its TelePresence immersive collaboration tools.

“Our goal has been to make those three prongs of the strategy work really well together, and we’re not there yet,” said de Castro. Over the next few months, Cisco will roll out a number of new solutions expected to help the trio gel.  

Two recent acquisitions nurtured Spark’s evolution and gave rise to the new Spark Assistant — an enterprise-ready voice assistant designed for meetings. In 2015, Ciso acquired Tropo Inc. to build out its developer ecosystem and boost its collaboration as a service business. Then last May, the company bought MindMeld Inc., an AI chat and voice platform.

“Spark is a digital assistant, just like Amazon Alexa or Apple Siri, but built for the enterprise with Cisco security behind it,” de Castro said.

Hard realities of enterprise AI

Cisco’s existing hardware presence in many companies will give it an easy in with customers, de Castro believes. “That is the single biggest asset that we have at Cisco in order to sort of penetrate this digital assistant, voice assistant market,” he said. “For us, it’s a software upgrade.”

Cisco does have devices with AI baked in, like its Room 70 all-in-one codec-quad, camera-70-inch single or dual display device with integrated speakers and microphones. The Cisco Spark Board for wireless presentation, digital whiteboarding and video conferencing was built with Nvidia Corp.’s Jetson TX1 visual and AI computing platform that also powers robots and drones.

The hardware question is mapping the AI assistant market in interesting ways. Some speculate that Amazon.com Inc.’s lack of a smartphone will cost it its lead. It’s Alexa-powered Echo device now has about 71 percent of smart-speaker market share, according to analysts. But Alexa does not have a smartphone built for it like Google Assistant (Pixel phones) and Bixby 2.0 (Samsung Galaxy phones), for instance. The Alexa app is downloadable, but this seems a bit counterintuitive for users of phones with an assistant already built in.

Cisco is aiming specifically for enterprise AI, not individual consumers. While Amazon introduced Alexa for Business recently, de Castro believes that Cisco has the upper hand against it and other competitors. “We already have the hardware in place,” he said. “For some of these other companies to kind of get into the conference room, they first have to convince IT, facilities and everyone to kind of install this new thing that is an unknown quantity.”

Training day for AI hopefuls

Cisco isn’t leaning entirely on the hardware install base as a crutch, however. It understands how crucial advanced AI software models are, and that they require much fine-tuning and massive data sets, de Castro explained. “Now we’re looking at, what are the data sets that we have at Cisco that we can use to allow our customers to derive insights from the collaboration that’s taking place that no one else can do?” 

In fact, any product in the AI market stands or falls on the data that train its models. “Training makes all the difference in the world between whether a predictive analytics ML algorithm actually predicts what it’s supposed to predict or doesn’t,” said James Kobielus, Wikibon.com’s lead analyst for data science, deep learning and application development.

It turns out, Cisco’s data from collaborations and meetings can reinforce the accuracy of its AI assistant. “Cisco’s sitting on literally billions of minutes of video and audio recordings that we can be doing a lot with,” de Castro said.

It is applying machine learning, speech recognition, speaker and meeting summarization, and natural language processing to surface insights from the data. MindMeld came to Cisco packaged with a platform called Workbench, which can derive semantic insights from text with NLP. Such insights could be useful to members of meetings or executives, like chief information officers, to gauge information about the meeting and its members, de Castro pointed out.

Spark is aiding Cisco in its journey to being a platform with its open application program interfaces. Cisco has taken out nuggets of the Spark application and allows people to embed them in third-party applications like Salesforce. Companies can have a full-featured Spark experience without ever leaving Salesforce, de Castro stated. “No one else can say that,” he said.

De Castro has pushed Cicso toward openness as a platform since he joined the company. “I said, ‘Look […] Cisco has an opportunity to be the AWS of collaboration,” he said. While it doesn’t quite have the muscle yet, all of the ingredients are there, and some customers are already consuming its services without using its actual apps, he added.

“That’s a pure platform play,” de Castro concluded.

Watch the complete video interview below, and be sure to check out more of SiliconANGLE’s and theCUBE’s coverage of Cisco Live Barcelona 2018. (* Disclosure: Cisco Systems Inc. sponsored this segment of theCUBE. Neither Could Cisco Systems nor other sponsors have editorial control over content on theCUBE or SiliconANGLE.)

Photo: SiliconANGLE

A message from John Furrier, co-founder of SiliconANGLE:

Your vote of support is important to us and it helps us keep the content FREE.

One click below supports our mission to provide free, deep, and relevant content.  

Join our community on YouTube

Join the community that includes more than 15,000 #CubeAlumni experts, including Amazon.com CEO Andy Jassy, Dell Technologies founder and CEO Michael Dell, Intel CEO Pat Gelsinger, and many more luminaries and experts.

“TheCUBE is an important partner to the industry. You guys really are a part of our events and we really appreciate you coming and I know people appreciate the content you create as well” – Andy Jassy

THANK YOU