How telcos can boost business by implementing AIOps

By: Eric Walker

As the quest for continuous connectivity drives change across the telecommunications industry, one thing has become clear: AIOps is more than just another IT solution — it’s a fundamental enabler of transformation and growth.

Heavy Reading’s 2024 5G AIOps Operator Survey, commissioned in part by Kyndryl and Microsoft, found that more than a third of telco leaders expect AI to be fully integrated into their operational processes within two years.¹ However, only 6% currently use AIOPs to identify and resolve all network issues.

For a sector facing shrinking margins, stifling competition and increasing regulatory pressure, there’s plenty of incentive to close the gap between AI ambition and actual implementation.

Will AIOps soon be non-negotiable for telcos?

The telecommunications industry is moving toward a future dominated by 6G networks, digital twins, cloud computing and other emerging technologies.

Since traditional networks and management tools aren’t designed to handle these advanced applications and services, IT complexity will grow exponentially. Telcos that don’t adapt will be unable to maximize the potential of new technology and meet customer demands for high-quality, reliable service.

On the other hand, CSPs that use AIOps to monitor and manage their IT operations will be better positioned to thrive. By combining AI, machine learning and big data, telcos can automate many IT operations processes, including event correlation, anomaly detection and causality determination.

From a business perspective, AIOps should play an integral role as telecommunications companies try to remain relevant. Research from McKinsey shows that telcos which incorporate AI technology as part of a holistic transformation can increase revenue by up to 8% annually, reduce costs by 10% to 15%, and boost customer satisfaction scores by 20 to 40 points.²

AIOps use cases for telcos

Some of the most impactful applications of AIOps in telecommunications would enable seamless technology integration, allowing telcos to quickly adjust to changing market demands. The potential use cases include:

• Predictive problem-solving. Engineers can use machine learning algorithms to analyze historical data and current network conditions to forecast issues and potential failures before they happen, reducing downtime and service interruptions.³
• Automated incident management. Chatbots, virtual assistants and other AI-driven automation tools can help telco customers troubleshoot and resolve routine inquiries, which allows customer service agents to focus on more complex issues.⁴
• Network optimization. AIOps facilitates continuous network monitoring and observability, enabling IT personnel to allocate additional network bandwidth or re-route traffic in real time to maintain optimal performance during peak hours.
• Security and resiliency enhancement. AI-based systems help detect and mitigate cyber threats by identifying unusual network activity and initiating security protocols to protect the network or speed restoration if there’s a cyber incident.

How to overcome barriers to adoption

While companies in other industries have fast-tracked the integration of AIOps into their systems and processes, many telcos have been slow to embrace the practice. Here’s how CSPs can accelerate implementation by addressing five of the main barriers to adoption:

1. Barrier to adoption: outdated architecture

Many telcos rely on systems and processes developed over decades to run their IT, making it difficult to integrate modern AI solutions without disrupting current operations. Outdated architecture — especially among carriers that haven’t transitioned to cloud-native operating models — also struggles to handle the vast amounts of data generated by modern networks or process information fast enough for real-time insights.

How to address: Commit to modernizing your infrastructure in phases, using cloud-based platforms and APIs to streamline integration with AIOps tools. You can also minimize disruptions by upgrading existing hardware, network bandwidth and data processing algorithms to handle the added demands of AIOps.

2. Barrier to adoption: data silos

CSPs generate information like performance metrics, customer interactions and service logs from various sources, so data often gets fragmented and stored in disparate systems. These data silos inhibit aggregation and analysis, which makes it harder for engineers to continuously monitor operations and make decisions in real time.

How to address: You can amplify AIOps by consolidating information from across your organization into a central repository like a data lake, where it can be accessed and analyzed more efficiently. You’ll also want to prioritize data cleansing and validation to ensure the information that powers your AIOps systems yields accurate insights. Data integration solutions that provide a unified view of your IT operations can be particularly beneficial in consolidating and managing data.

3. Barrier to adoption: security and privacy issues

Adhering to data protection regulations like GDPR, HIPAA and CCPA — coupled with integrating AIOps with existing security tools and frameworks — is complex and time-consuming. And even though AIOps can enhance threat detection capabilities, any weaknesses in AI algorithms or data processing pipelines create potential vulnerabilities for cyberattackers to exploit.

How to address: If you haven’t already introduced a zero-trust policy, now’s the time to act. For enhanced security, you can implement encryption tools, anonymization techniques and continuous monitoring. You’ll also need to put governance policies in place to safeguard data and ensure compliance with all industry standards and regulations.

4. Barrier to adoption: talent shortages

Like countless other companies, telcos often struggle to find and retain employees with the expertise in AI, machine learning and data analytics needed to oversee a successful AIOps program. These skills gaps can slow implementation, increase costs and, potentially, reduce the effectiveness and ongoing reliability of AIOps solutions.

How to address: Upskilling and reskilling your workforce is a fast, effective way to help employees gain the knowledge they need to implement and run your AIOps practices. You can also work with vendors and third-party experts for role-specific training and establish internal Centers of Excellence to support ongoing training and share best practices with employees. If necessary, consider partnering with systems integrators or outsourcing firms to build an AIOps team.

5. Barrier to adoption: resistance to change

Whether it’s fear of losing their job, lack of understanding or skepticism about AI benefits, employees may be reluctant to embrace AIOps. However, resistance to change — especially among leaders — can prevent telcos from reaping the full benefits of the practice.

How to address: Communicate early and often with employees, underscoring how AI-enabled work can enhance their jobs rather than replace them. Executives should also illustrate how AIOps adoption aligns with your organization’s overall business strategy and promote this vision at all levels of the company. Once executive sponsorship is in place, you can appoint advocates to champion AIOps adoption among coworkers, highlighting the benefits it can deliver.

A competitive advantage for the future

The disconnect between AI aspiration and application will create a competitive disadvantage for telcos that don’t adapt. By addressing common barriers to adoption and implementing AIOps, telcos can build the operational foundation needed to thrive in an AI-driven future.

Eric Walker is director of strategy for Kyndryl’s U.S. technology, media and telecommunications market.

¹ Heavy Reading’s 2024 5G AIOps Operator Survey, LightReading, September 2024
² The AI-native telco: Radical transformation to thrive in turbulent times, McKinsey & Company, February 2023
³ Harnessing the power of AI in telecommunications, Forbes, July 2024
⁴ What does 2024 have in store for telecoms?, Telecoms.com, December 2023