Despite the hype, AI is real and provides the foundation for supporting a rapidly changing digital transformation environment. In the span of a few years, the availability of low-cost cloud computing with high-performing computing technologies such as Graphical Processing Units (GPUs) and Tensor Processing Units (TPUs) drove rapid AI innovation.
To broaden adoption, AI technology vendors added model development tool kits, pre-trained model templates, and educational services to make it easier for companies without large teams of data scientists to develop and operate AI models. These efforts enabled AI to evolve from the theme of science fiction movies to a set of solutions used daily by millions of people in their personal and professional lives. Networking vendors are also embracing AI to deliver intelligent networks.
The right networking technology allows IT to deliver what Lopez Research calls Right-time Experiences. A networking RTE is a set of infrastructure and solutions that provides IT leaders with the right insight at the right time. RTEs require a networking infrastructure that can deliver services that are contextual, adaptive, learning and predictive.
To create these predictive systems, leading networking vendors combine machine learning and deep learning with their data corpus to create useful out-of-the-box AI-enhanced solutions. AI helps administrators collect and analyze data from various sources to anticipate issues, enable automation of specific tasks and optimize overall network performance.
This field is often called AIOps. For example, AIOps helps a network administrator efficiently troubleshoot issues by analyzing large amounts of network data, identifying a problem, and surfacing what alerts require attention. AIOps eliminates alert fatigue. But more than that, the network service goes beyond outlining what's wrong. It helps IT figure out what to do next.
AIOps is a human-assistive technology that enables administrators to work on higher-value tasks by automating repetitive tasks such as configuration management, RF optimization and troubleshooting. It allows IT to be more strategic by delivering advanced networks. Successful IT leaders will focus on working with the business to understand how the network can support new business goals. Companies have completed the first round of digital transformation. It's time to use AI to move on to what's next.
What to look for in a vendor's portfolio
One of the challenges with AI is that every technology vendor claims its product is AI-powered. These claims frequently led to scenarios where tech vendors overpromised and underdelivered in AI. In some ways, AI was overhyped as a technology that would solve anything. Watching customers struggle with this issue has led me to conclude that companies should be evaluating what type of AIOps their networking vendor offers and how to deploy these services effectively.
Ideally, your vendor would offer a console that unifies network provisioning, monitoring and troubleshooting. AIOps should bridge the cloud and on-premises environments and supports evaluating performance from the app through the network. In addition to troubleshooting, AIOps should also help IT detect patterns to optimize the network and support new applications. AIOps allows IT to move from reactive to proactive by pinpointing the root cause of current issues, predicting potential issues and suggesting the next best actions.
IT leaders need to evaluate their technology vendor's approach to the AI lifecycle, spanning from design to set up through operations and refinement. The solution's efficacy starts with data because an AI solution is only as good as the data that creates the models. When assessing AIOps, it's essential to work with a company that has a good data foundation and domain-specific data to develop suitable models. Typically, this leads us to more established providers with a large corpus of network data and insights to analyze.
Next, a company should evaluate how easy a networking AI system is to set up. IT should look for solutions that provide integrated and adaptive management. Ideally, AI would be a feature that is easy to turn on with updated software from your technology vendor. Third, you should evaluate the operations, automation and controls functions. What does it allow you to automate? What are the controls in place? Finally, your technology vendor should provide you with a roadmap for AI updates because the AI field is constantly evolving. Your technology partner should be continually refining its solutions.
What's beyond AIOps for troubleshooting network issues?
Troubleshooting is where you start, not where you end. AI is a game-changer for networking. AIOps help IT gain control through a better understanding of the network. As a result, IT can redeploy its resources to focus on tasks that provide a higher return on investment than managing basic network functions.
After troubleshooting, most companies use AIOps to improve overall network performance by having the system recommend configuration changes and revisions to application policies and settings. AIOps can help IT proactively predict when and how the company should upgrade its network. Eventually, there's the opportunity for AI to provide closed-loop automation, which helps us realize that vision of a self-driving network for specific aspects such as onboarding and securing new devices.
Overall, AI-powered networks allow IT to meet the increasing demands for network performance and enable new experiences such as adding augmented and virtual reality applications for designing facilities and repair, enabling remote robotic process control in factories, and supporting video analytics in retail and warehouses.
What excites me the most about AI is that it is a grand equalizer. Companies, regardless of their size, can take advantage of AI in networking products today. The only question now is how your company will use intelligent networks to advance your business.
