The next generation of mobile networks is around the corner and Guavus CEO Faizel Lakhani is fired up about it. Helping telecom operators keep pace with 5G changes is one of several ways Guavus can help as it grabs millions of bits of network data, puts it instantly into context and helps the network itself predict and prescribe what happens next.

"5G's coming along and you're seeing huge, plentiful data," he said, pacing back and forth behind his desk as we chatted via video conference. "The problem is 5G and NFV and SDN are kind of woven together because as the network capacities explode, then the ability to have dynamicism and variability is high."

Indeed, he actually said "dynamicism," a possible combo of dynamism (vigorous activity) and dynamic (constant change). Given Lakhani's enthusiasm, it seemed to fit.

His point is that all of these major things about telecom networks -- the way customers use the network, the topology of the network, the software powering the network -- used to be fairly static. It's why, back in the day, he was among a loud chorus of industry companies pushing telecom operators to do more than just provide connectivity. To keep growing -- to generate more network use and more revenue -- they were going to have to get creative. "People who are only selling pipes are the ones who will die," said Lakhani in 2001, who was then the vice president of network solutions at core router startup Caspian Networks. (See Report Heralds IP Doomsday and Caspian Closes Its Doors.)

While Lakhani's career moved him into enterprise networking and security software, following stints at Caspian and Nortel, the world changed. Telecom networks -- thanks to Facebook, Apple, Google and Netflix -- are riding a growing wave of network traffic. The massive bandwidth and low latency connections promised by 5G are going to upend telecom operations if the networks can't automate their processes and move quickly enough to satisfy demand and guarantee the delivery of services. That's why, Lakhani said, his company is doing so well right now.

Guavus grows up
Guavus was founded in 2006 as a platform to help companies make sense of unstructured data. It was bought by French conglomerate Thales in April 2017 for about $215 million and Lakhani was added as CEO in October. In December, Thales organized its digital assets and acquisitions into a single business unit, with Lakhani leading Guavus and reporting to the unit's head, David Jones, who was formerly the security and governance unit of HPE. (See Thales Splashes $215M on Guavus for Analytics and Liberty Global Invests in Guavus.)

Guavus won't disclose its revenues or relative growth stats. It claims to have seven of the top ten North American cable operators and the five largest North American mobile operators as customers.

Against that backdrop, Lakhani, the company's third CEO in five years, is managing an organization with a mature big data platform as it layers on artificial intelligence capabilities and aims to find new business cases for operators to use to shave costs off every bit they deliver -- even as the number of those bits is about to skyrocket yet again. Lakhani adds: "These opportunities -- I'm talking about 5G, NFV... the desire by carriers to reduce customer care costs and move to a Google-like customer experience model... these are new drivers that are really increasing the adoption of our solutions. Because you can't do it the same old way that you used to do it, right? It doesn't work." (See Guavus Gets a New Top Dog.)

Figure 1:

Guavus CEO Faizel Lakhani, corporate headshot.

Guavus was getting attention by helping service providers pull network data, CPE data, info coming from routers, phones and other network endpoints and create sophisticated reporting on all of that big data collected and stored. But, Lakhani said, service providers needed more. "That's where the AI piece comes in... it's all about that automated action" based on the real-time analytics Guavus was already good at, he said.

"How do you know what you should do and when things are going to happen," Lakhani said. Solve that -- and do it automatically -- and you'll reduce the costs of running a service provider network. Depending on the application, Lakhani said Guavus is helping service provider automate operations, make better use of their resources and giving them ways to let customers take control of how services are delivered.

Figure 2:

CEO Faizel Lakhani, outside his office.

Now Guavus is reducing truck rolls by giving customer service reps in some service providers a thorough enough dashboard on each customer that they can quickly take prescribed actions. Rather than endlessly quizzing customers, the Guavus AI engine looks at all the incoming data about the caller, their devices, their service history and makes recommendations. "We saved that [service provider] customer $70 million a year -- seven, zero -- in the elimination of truck rolls because we're augmenting their human call agents with an AI system that can help them remediate," Lakhani said.

The next step, he said, is to use network data to prevent the calls in the first place, perhaps by pushing a firmware update to a router well before a software incompatibility slows the consumer down.

AI has an eye on you
Another Guavus push for service providers is to help them reach consumers with offers at the exact right moment that they're ready to buy.

AI technology can help predict "when the customer's not going to have a good experience or when you're gonna have a network outage," he said. Or, it could tell the service provider, "'Hey, this person has an intent to buy something. This person is looking for a car.' It's that intent and prediction that is really where our focus is to come," he said.

Lakhani said Guavus can help a service provider correlate a customer's age, location, ad responses, their shopping tendencies and merchant preferences, the apps they use, the websites they visit and all kinds of other info service providers can pull from disparate parts of their network and devices.

Put more simply: Guavus can give service providers Facebook-like capabilities to know a customer's every move.

"The interesting thing here is that you've got service providers that actually have access to a ton of the data," he said. "They also know location and things like that and, from a Guavus perspective, there's a number of technologies, that are AI-based, that help a service provider monetize the pipe."

Lakhani said Guavus' AI engine can build a picture of a group of customers in an abstracted report and fairly accurately predict what they'll do next. "You would be able to say, 'Hey, these particular users have this type of intent,' and then you combine that with other information like demographics, and you can do some really, really interesting things around that."

We'll see just how far Guavus' customers go with AI-based consumer predictions as GDPR kicks in and consumer privacy concerns are hitting an all-time high. What hasn't changed is Lakhani's ability to look beyond the here and now.

Yes, 5G is coming. Yes, networks need optimizing and they need to respond more quickly than ever to the demands consumers will put on them.

But Lakhani won't stop there. Like he did 17 years ago, Lakhani can make a case that service providers can use their networks to do more than just deliver bandwidth. Before, he was pushing the idea of helping them more wisely sort their traffic. Now, he's giving them the tools to predict what you do next.

"Business is awesome," he said, finally sitting back down. "It's actually a good place to be right now."

— Phil Harvey, US News Editor, Light Reading