Jonathan Forest, VP Analyst, Gartner, Inc.February 7, 20255 Min Readestherspoon via Adobe StockAI networking has been a hot topic over the past few years and is a subset of AIOps. Generative AI (GenAI), which is part of AI networking, has taken this hype to a new level with the potential to transform network operations. However, with its conversational interface and ongoing learning capabilities, GenAI will likely be met with both favor and distrust.But what can enterprises really gain by using GenAI as part of the network operations? CIOs must be aware of new GenAI capabilities for network operations, business case considerations and ways to build trust to minimize adoption risk.GenAI promises great potential to enable improvements to long-standing traditional networking operations practices across Day 0, Day 1, and Day 2. With GenAI, network operations can accelerate initial configurations, improve the ability to change vendors, drive more efficient troubleshooting and simplify documentation access.Day 0For Day 0, for example, an engineer could use an iterative process and ask the GenAI network tool via a natural language interface to design a leaf-spine network to support 400 physical servers using Vendor X. Additional information like SLA requirements (such as availability and throughput) can also be included via natural language to deliver the desired performance level and design that includes cost implications.Related:Another example is in the area of capacity planning as new users, applications, and architectures are adopted, making network planning more complicated. GenAI can be used to help size network infrastructure and optimize costs based on the number and types of applications hosted on-premises, in the cloud and at end-user locations (in the office, at home or other locations).Day 1The GenAI network tool can then help generate/validate/optimize all the required Day 1 configurations based on desired criteria (for example, by price or performance). It may not be 100% accurate, which is why it may require an iterative process to refine GenAI tool outputs to accelerate/optimize network setup. Even if it requires several iterations, the use of GenAI would represent a substantial improvement over current rigid processes and tools, reducing time and errors by up to 25%. We envision that this will be leveraged in all networking domains (WAN, data center, cloud, and campus) to assist in the design and setup of networks.Day 2AI networking enhances Day 2 network operational support by correlating multiple data inputs, identifying problems faster, yielding quicker resolution and, where applicable, spotting problems proactively before a user is aware. GenAI will bring additional capabilities including a conversational interface and the ability to learn over time. It can also enhance user experience with specific outputs such as text, audio, video, or graphics.Related:For example, to help isolate problems, CIOs can ask GenAI to build a dynamic graphic of networking performance issues over time based on packet loss, latency and jitter. It can also focus on specific questions such as Is the CEO having network performance issues?GenAI can create detailed configurations and troubleshooting procedures based on natural language inputs without explicit templates. GenAI tools can drive network operational support time savings by up to 25% when compared with the status quo by driving efficiencies that cant reasonably be achieved by scaling manual resources. It removes manual processes to identify issues more quickly, resulting in faster problem resolution.Calculate the Value Before InvestingCIOs must ask pertinent questions to gain a complete understanding of the inherent value of GenAI networking, its use cases and common tools. A key facet in the process of GenAI adoption involves building the business case and calculating the value to the organization.Related:Asking pertinent questions can offer more insights while creating a business case to determine the value of GenAI functionality. Specifically, determine if aligning network operations with GenAI can help build scale, control/reduce costs, drive resource efficiency, foster agility to keep up with the digital business and deliver a better end-user experience.Prove the ConceptFirstIn addition to the immaturity of GenAI networking functionality and the need to quantify the value, another key limitation that needs to be overcome to achieve wider adoption by network operations is a lack of trust. Network teams have been burned many times by vendor claims of automation or single pane of glass to solve existing issues. This, in part, is the reason why network operations teams have been slow to adopt network automation and are skeptical about GenAI. On top of this, GenAI networking tools may yield inconsistent responses, which introduces risk and fosters mistrust.However, network operations teams need to include GenAI functionality in their RFPs/RFIs to determine the scope, value and capabilities of the solutions in the market as they mature.Running a proof of concept (POC) is key for network operations personnel to determine the accuracy of the GenAI solution, alongside its maturity, level of trust and degree of comfort. This is really more about quantifying the accuracy of the GenAI networking solution across a wide range of scenarios. Even in production, we expect network operations personnel to have to validate some or many GenAI outputs, but baselining the capability gives context to the accuracy and the level of unsupervised trust (if any) that should be given.When running the POC, begin by testing in a lab environment before moving to a real-life production environment. Test the solution over several weeks and months to stress it as much as possible. Have multiple personnel leverage the tool to capture multiple opinions/perspectives. Validate the GenAI networking tool outputs for accuracy by testing against alternative sources. Measure the time to perform tasks with the GenAI networking tool and with the previous/current method. In short, the goal is to compare process efficiency and accuracy of the current approach versus the intended GenAI approach. As part of this POC, both the level of trust and value (business case) can be determined to help inform a sourcing decision and simplify adoption, if applicable.If CIOs follow these suggestions, they will uncover the value that GenAI can bring to network operations while also gaining trust.Read more about:Network ComputingAbout the AuthorJonathan ForestVP Analyst, Gartner, Inc.Jonathan Forest is an Analyst within Gartner's Information Technology Leaders team, focusing on Infrastructure and Operations with respect to enterprise networking. Mr. Forest primarily performs research and advises on SD-WAN, SASE, AI networking, infrastructure to support AI, cloud networking, managed network services, application performance, LAN/WLAN, data center networking, network automation, NaaS and more.See more from Jonathan ForestNever Miss a Beat: Get a snapshot of the issues affecting the IT industry straight to your inbox.SIGN-UPYou May Also LikeWebinarsMore WebinarsReportsMore Reports