Cisco Uses AI to Bring Advanced Capabilities to the Network

NEWS ANALYSIS: At Cisco Live 2019, Cisco announces a number of advanced features to make the network smarter and more secure.

Cisco 1989

Queue up Kid Rock’s “All Summer Long”:

“It was 1989, my thoughts were short, my hair was long, caught somewhere between a boy and man.”

What’s this have to do with networking? In addition to “Singing Sweet home Alabama all summer long,” 1989 was the year that Cisco held its first-ever Networkers customer event. Only a few hundred people attended that event as Cisco’s customer base and product portfolio was significantly smaller than it is today.

Fast forward 30 years and not only has the name “Networkers” been retired in favor of the broader and more modern Cisco Live brand, but attendance has grown to over 28,000 as Cisco has gone from a niche router vendor to one of the world’s most dominant IT vendors.

Networking Is What Cisco Still Does Best

Its portfolio has evolved and leadership and strategy have changed, but at its heart Cisco is still doing what it’s always done better than any other company, and that’s drive innovation into the network.

It’s been interesting to be a Cisco watcher for the past 30 years as the company has faced a barrage of threats to its success and even existence: Market maturity, ProCurve, Huawei, white box, software-defined networks and the cloud were each viewed as the thing that would destroy Cisco’s business. Of course, none of these did that, and Cisco is still enjoying dominant market share and well-above-average industry margins.

The secret to its success isn’t really a secret at all. Cisco has managed to stave off commoditization and low-cost threats by constantly driving innovation into its products. I believe very few markets actually ever become commodities. Apple has kept the innovation curve high on laptops and phones, and its business has been solid for years. Cisco has delivered a steady drumbeat of new features that has enabled it to do more than a company built on competitive products. 

Cisco Is Advancing the Network With AI and ML

The network announcements Cisco made at Cisco Live 2019 last week were another example of innovation to deliver enhanced capabilities. However, these new capabilities had nothing to do with speed and feeds. Rather, they make the network smarter, simpler to manage and more secure through the use of artificial intelligence and machine learning (AI/ML). As part of its intent-based networking (IBN) strategy, Cisco announced software that enables IT teams to work faster through personalized network insights.

Complexity Is Killing IT

During his keynote address at Cisco Live, CEO Chuck Robbins mentioned that most businesses depend on their IT organization to transform digitally. The challenge for IT is that complexity has shot through the roof but headcount has remained relatively flat. IT needs to operate faster than ever in an environment that’s more complicated than ever. This is creating an untenable situation for IT pros.

AI and ML Enable Network Pros to See More and Secure Better

Cisco is doing its part to reverse this trend through the use of AI and ML. The company has a huge install base and gathers a massive amount of contextual telemetry data that holds many insights network operations can leverage for better security and simplified network operations.  Cisco is using is using AI/ML to analyze the network data to provide IT teams with the following:

  • Improved visibility: Data is collected from the local network, and baselines are created for devices, users and applications. As the environment changes, IT operations are immediately informed through a dashboard. Bring your own device (BYOD), internet of things (IoT), the cloud and other trends have created an environment in which most organizations have no idea of how many cloud apps are being used and what devices are connected. The baselines continually learn and adapt as things change.
  • Relevant insights: Cisco has collected data from across its customer base. It then uses AI to process the information and establish industry-wide baselines and norms. Company-specific data is then compared to the aggregated data to discover insights where the network could be improved. This can also help predict problems so network engineering can proactively fix issues before users are impacted. This can be particularly helpful for WiFi issues, which are very difficult to troubleshoot.
  • Guided Actions: Machine learning is used to automate workflows to perform the logical troubleshooting steps that an engineer would execute to resolve a problem. This helps IT detect issues and vulnerabilities, analyze the root cause and execute corrective actions faster than ever.

Multi-Domain Networking Reduces Complexity

At Cisco Live, the company also announced IBN for operational consistency across every domain of the network—campus, branch, WAN, IoT, data center and cloud. Customers can now create policy and management uniformity everywhere. The low-hanging fruit use case for this is network segmentation. Almost every network pro I talk with is interested in segmenting the network for security purposes. The problem is that WAN segmentation is typically done independently of the data center, which is done separately from the campus and so on. Through the integration of SD-Access, SD-WAN and ACI, segmentation policies can be created to consistently authorize, on-board and segment devices and users regardless of where they are.

Another use case for multi-domain networking is creating policies for application performance. Most applications, particularly cloud ones, traverse numerous domains, making traffic prioritization difficult. Cisco’s multi-domain lets this be done through a centralized console.

The pace of change and dynamic nature of IT are making it very difficult for businesses to transform digitally. Cisco is combining the power of AI/ML with the massive amount of data it collects to simplify network operations to improve application performance and make the environment more secure. Cisco Live 2019 was a big one as it once again demonstrates that innovation will always trump commoditization.

Zeus Kerravala is the founder and principal analyst with ZK Research. He spent 10 years at Yankee Group and prior to that held a number of corporate IT positions.