Beyond the Hype: Understanding the Real Value of AI for Service Providers

April 30th, 2024

I recently wrote about how AI can support unified communications in the enterprise, but its applications are much broader than that, and communications service providers also have the opportunity to benefit. In a recent Gartner survey, 95% of telecom CIOs said they plan to implement AI/ML technologies by 2026 and 79% said they plan to increase their AI/ML investments. These providers are struggling to drive expense out of their networks as they have yet to see meaningful returns on their massive 5G RAN and spectrum investments. Additionally, mobile ARPU is flat in many countries and some reports have even shown significant ARPU declines in places like the UK and Germany.

So where should CIOs be using AI? In my mind, the logical place to start is to use AI to automate network operations and reduce deployment and operational complexity. By leveraging AI /ML to simplify and automate tasks, service providers can lower OPEX and increase service velocity. That’s not to say AI can’t be a revenue generator, but those efforts require seed projects to test the waters. From a pure ROI perspective, AI is tailormade to deliver OPEX savings now.

Intelligent Application Lifecyle Management

Application lifecycle management is a great example of how AI can reduce OPEX. Service provider networks are never homogenous; they include hardware-based equipment, virtualized software elements, newer cloud-native network elements, etc. Testing and deploying new applications or software releases is typically still a manual process that’s resource-intensive, time-consuming, and disruptive. It can take months or even years to perform a major software upgrade. In the meantime, inherent security vulnerabilities can leave the door open for threat actors.

AI is great at crunching massive amounts of data from your network. AI-based tools can interrogate your network, identify edge cases, and automatically create and run test suites tailored to your specific environment. Even the best traditional test team can only validate a subset of all the uses cases in a large network. AI-powered testing expands test coverage and exposes flaws sooner. It also makes more efficient use of testing resources and frees up testers to dig into other issues.  By identifying and resolving issues earlier in the application lifecycle, you can accelerate deployments and upgrades, and have greater confidence new software will work right the first time it’s deployed.  

AI/ML-powered toolsets can keep working after the upgrade or deployment is complete.  As the network evolves the toolset can adapt test cases, automatically covering new call flows, network elements, and connections.

Intelligent Troubleshooting

In a moment of crisis, service providers still mostly rely on the raw talent of their teams to identify and isolate service quality issues, performance bottlenecks, and security breaches. It can take even the most experienced support engineer hours or days to wade through hundreds of alarms and thousands of datapoints to pinpoint a complex problem spanning multiple systems and network technologies. Plus, as humans we often leverage our vast experience to solve problems. That serves us well when an issue repeats itself, but it may derail us if a new problem has similar symptoms to a familiar issue. 

AI isn’t going to replace your best technical minds, but it can certainly assist them.  It can also be impactful at 2 AM when those great minds aren’t available (or aren’t thinking clearly).  AI-powered troubleshooting tools can instantly analyze vast amounts of end-to-end data, up-and-down the entire network stack, to identify the most probable cause of an issue. They can save hours and prevent a technician from going down the wrong path.

Speaking of 2 AM, AI makes it easier for less-skilled staff to manage complex issues. Intelligent troubleshooting assistants can interpret spoken or written language, so technicians no longer need to know 20 different CLIs. They can also translate cryptic system alarms into plain language that less-experienced staff can understand. Intelligent troubleshooting assistants improve productivity for both seasoned employees and new hires.

Predictive Analytics

AI is great for identifying and resolving potential network issues before they impact customers. There can be many smoldering problems within a network which only rise to the surface at the time of failure. AI can analyze the network and highlight potential risks prior to the failure occurring. Along the same lines, networks can be pretty “noisy.” Understanding which noise is “normal” and which is a sign of bigger problems is an important use for AI. Intelligent analysis and predictive analytics can save massive amounts of time and money, as well as minimize customer experience problems.

Not Just Cookie Cutter Options

I suspect you may be thinking, “that sounds great, but you have no idea how convoluted my network is.”  Fair enough, I don’t. To be successful, these new AI tools must be easy to develop and easy to customize. From a Ribbon perspective, we are building automation platforms that let service providers connect multiple network elements to create unique workflows, using low-code/no-code development.  So, you can observe, test, and automate your network based on your unique environment and your goals. 

Low-code/no-code development platforms make it dramatically easier to build and customize automation scripts and applications – without a “lifetime” pro services engagement. In fact, in the Gartner survey referenced above, 91% of telecom CIOs said they plan to introduce low-code/no-code development platforms within two years. They can also help you speed up AI initiatives and accelerate time-to-value.


These are just three basic examples of how AI and ML can help you accelerate business agility, streamline operations, and potentially save millions. There are plenty of other practical applications for AI/ML like network performance and cost optimization, prescriptive maintenance, and intelligent threat detection and remediation. By focusing on pragmatic applications, you can improve business results quickly and make the most of your AI investments without getting caught up in all the AI hype.