The race to deploy AI and to become ‘AI-native’ is no longer an option but a necessity. It’s not about proofs of concept or ‘AI-enhancing’ existing services, but instead, a new strategy for building future-ready communications networks. Being bold and putting AI at the core of networks and operations to fundamentally reimagine services will unlock unprecedented efficiency gains whilst enabling vital diversification. The good news is, done well, AI can and should cement the telcos’ indispensable role in the ongoing digital transformation of our world.
Not just gen AI
Today’s telecommunications networks are vast, intricate ecosystems, characterized by what Analysys Mason calls inherent complexity and substantial operational costs. The sheer scale of provisioning, maintaining, and continually optimizing these networks represents a significant proportion of the resources, time and expenditure of a telco. This is, of course, the ‘day job’ of communications service providers. But, to what extent can these vital tasks be automated and transformed? Clearly, much of this critical work is too sensitive to trust to Large Language Model (LLM)-based AI with its (albeit reducing) propensity to hallucinate.
One approach is the Global Telecom AI Alliance (GTTA) formed by SK Telecom, Deutsche Telekom, e& Group, Singtel and Softbank. It is looking to co-develop a multilingual LLM to identify and solve telco-specific business issues using of local languages, plus industry-specific terms and their own internal business jargon. Delivering this, with the level of personalisation and automation required, would be exceptionally difficult with generic LLMs yet will still require significant time and investment from the Alliance.
But it is important to see beyond the current hype surrounding generative AI.
Gen AI certainly offers exciting possibilities, but non-generative or predictive AI will continue to play a central role in core network operations. By leveraging predictive AI and machine learning algorithms, trained with well prepared, consistent and accurate network data, operators can automate provisioning processes, proactively identify and resolve network faults, and dynamically optimize network resources in real-time. This is already the situation in many cutting-edge telcos, and where Teradata has deep heritage.
Connecting unseen dots
The exciting next step is to train AI and allow it to undertake inference across diverse and extensive datasets at the heart of the business. Often these remain siloed and jealously guarded by respective departments, business units or even different organizations. However, breaking down these silos can allow AI to uncover unforeseen and unexpected opportunities. Imagine AI systems correlating seemingly disparate data points from network performance, customer behaviour and profiles, and even external environmental factors to predict potential bottlenecks or proactively suggest network enhancements that would otherwise remain invisible to human analysis but nevertheless negatively impact customer experience. This capability extends beyond reactive problem-solving to proactive value creation, enabling more resilient, efficient, and cost-effective network management.
For example, AI can more effectively manage and dynamically optimize virtualized network resources, allowing telcos to offer guaranteed performance and customized connectivity solutions for specific applications, including emerging opportunities such as mission-critical IoT deployments or low-latency edge computing services. AI-powered network slicing and connectivity as a service (CaaS) offers can become more tailored giving customers more choice whilst creating differentiation and increased margin.
Expanding your palette
Beyond connectivity, AI -Native telcos can play an enhanced role as partners providing a wider palette of services currently the preserve of Big Tech and specialist firms. Security and cyber risk mitigation services are just one example. The increasing, and likely long-term demand for cyber-threat detection and automated incident response offers AI-powered telcos the chance to expand their enterprise customer product portfolio.
The AI-native network will be a platform for growth; delivering differentiated and high-value services that extend far beyond traditional connectivity. But the journey is not without its challenges. Robust data platforms that ensure quality and availability of data are crucial to the development of the AI-native network. The more effectively data can be integrated across the organization, the more impact AI can have. This means not only dissolving silos between customer data and vendor-agnostic network data, but also across traditionally separate systems like business support systems (BSS) and operations support systems (OSS), as well as integrating data from emerging sources like IoT and industry-specific applications.
Creating the right mindset
This will be scary for many CIOs and individual business unit leaders, but creating a holistic view of all the data available to the telco will support AI applications that unlock new value by transforming rather than simply enhancing services. Teradata can help overcome these fears in several ways. We have an unrivalled track record of working with telcos to maximise the value of their data, whilst ensuring its quality, security and reliability. Teradata has the technology to bring AI to the data in clouds, hybrid-clouds or on-premises, minimising expensive and risky data movements, whilst also providing the speed and scalability that ensures AI models can run in enterprise-wide live deployments. This means you have the tools to run analytics in database to deliver rapid, cost-effective AI deployments that deliver immediate value to frontline network and operations roles.
Leaders across the sector also need to work to cultivate the right mindset to make use of these capabilities within their organizations. The AI-native telco will exhibit a mindset that is not only data-centric, but which embraces experimentation and fosters close collaboration between business and development teams. It will also focus on productivity and scalability of AI efforts, favouring models and data features that can be reused and rapidly deployed over compartmentalised, stovepipe developments.
Transformative intelligence
With these elements in place, today’s telcos are well positioned to generate profitable growth with AI at the heart of their networks. AI, fed on high quality data from across all areas of the organisation can drive flexibility, resilience and productivity. By combining insights gleaned from data captured across network, OSS, BSS and from every corner of the organisation, AI can help spot, catalyse and act upon new opportunities and deliver new value.
The future of the telco industry lies in placing the AI-native paradigm at its very heart, and working with trusted providers to ensure safe, scalable and efficient access to the data it needs to deliver breakthrough transformation.