Cisco says its predictive analytics engine knows when your network will fail

Cisco says AI capitalism no future predictions
Image credit: Wavebreakmedia

ITEM: Cisco has pulled back the curtain on its predictive analytics engine for enterprise networks, which leverages AI and ML to crunch piles of historical network data in order to spot and fix network failures and other problems before they happen.

The idea of predictive analytics isn’t new, of course – plenty of enterprises have been using it for a couple of years for apps like analyzing buyer bhavior, equipment maintenance and fraud detection. Virtual assistants like Siri and Alexa use it too. NTT Global use predictive analytics for its enterprise network offerings to “identify and resolve issues before they impact operational performance”.

Cisco – which has been publicly working on predictive software engines for the past two years – is essentially announcing that it’s going to start integrating its predictive analytics engine across its product line in “easy-to-use SaaS offers”, which is potentially big news for its enterprise customers.

The big reveal is more of a teaser than a detailed explanation. From the press release:

Cisco predictive networks work by gathering data from a myriad of telemetry sources. Once integrated, it learns the patterns using a variety of models and begins to predict user experience issues, providing problem solving options. Customers can decide how far and wide they want to connect the engine throughout the network, giving them flexible options to expand as they need.

That’s it. But you get the idea. And to be fair, the announcement on Wednesday seems designed as a taster for its upcoming Cisco Live event next month, where the company plans to make some more concrete announcements about which of its products will have integrated predictive capabilities.

While the obvious application for predictive analytics is preventing (or at least minimizing) costly network downtime – which a Cisco survey says is the top network challenge for 45% of IT leaders – Cisco also puts a lot of emphasis on what predictive networks mean for the customer experience, which has become a critical market differentiator:

People and businesses use and rely on applications for just about everything, and often an app is the critical first impression for customers. 57% of people say brands have one shot to impress them and that if their digital service does not perform, they won’t use them again. To deliver on the full promise of digital business, the industry needs a way to better predict network issues, proactively avoid issues, and ensure the best possible experience.

Cisco does run with the experience angle a bit too far, saying that when a predictive network prioritizes experience, “not only does it function like the human mind, but it is also created for the human mind.”

Well … lol no.

But it’s certainly true that the customer experience is the KPI that matters most in the digital economy (and yes, okay, the Web3 metaverse), so anything that helps meet that requirement is welcome.

As for what makes Cisco’s offering different, it’s hard to say right now. Cisco makes reference to the fact that (1) the engine draws from several advanced statistical models (The Register reports that Cisco has also experimented with “with long short-term memory and convolutional neural networks”), and (2) it has “enormous amounts” of historical network data with which it has put the engine through its paces.

If nothing else, Cisco has customer testimony to back up its claims. Cisco chief Chuck Robbins said in a statement that its predictive networks research was tested and developed with customers, and that early adopters are seeing “major benefits saving them time and money”.

Reuters adds that 15 customers were involved with the testing, including Phillips 66, Schneider Electric and The Adecco Group, all of whom have provided testimonial blurbs on Cisco’s website.

More info here and here.

Be the first to comment

What do you think?

This site uses Akismet to reduce spam. Learn how your comment data is processed.