- Axiata Group operates several service providers in Southeast and South Asia, including Dialog, XLSmart, Celcomdigi and Robi
- When combined, the group has over 175 million subscribers
- The company's group chief business and technology offers believes AI is likely to help it improve its financials by as much as 20-30%
Axiata Group’s focus on transforming its network infrastructure and operations through the implementation of autonomous networks (ANs) and AI is likely to help it improve its financials by as much as 20-30%, believes Thomas Hundt, group chief business and technology officer at Axiata Group.
“We have already done a great deal over the last couple of years on CapEx optimization and maximizing the value of existing assets through intelligent network planning. I believe that when network resources are augmented with AI in a more intent-based fashion, aligned to whatever the user needs, we will be able to deliver the right resource at the right time. We see an incremental potential of 20-30% still available in terms of network uptime and resource optimization,” Hundt told Fierce Network on the sidelines of FutureNet Asia 2025.
Axiata Group is a prominent telecom conglomerate in Southeast and South Asia. It operates several service providers in the region, including Dialog, XLSmart, Celcomdigi and Robi. Together, it has over 175 million subscribers in the region.
The group shared its A3 (Axiata AN and AI) strategy last December to transform its network infrastructure and operations through the implementation of ANs and AI. The key features of the A3 Network Architecture are openness, programmability, cost efficiency, AI native, TMF/Open Digital Architecture (ODA) conformity and scalability.
“The A3 architecture for us is the blueprint of how our network architecture should look, modular openness embedded, service orchestration layer and SMO [service management and orchestration] with our apps orchestrating the network. We are rolling out the plan across the group as SMOs by the end of next year and then to leverage our app ecosystem,” Hundt explained.
Axiata launched its enterprise transformation program, Axiata AI, a little over a year ago, where it started building the foundations of AI native telco. “It has four pillars: people, infrastructure underneath, which is data architecture and then AI factory, where we build our use cases — we have around 50 use cases now in operations. And the fourth is governance, which is non-negotiable.”
Anomaly detection, energy savings and digital twins for towers at scale are some of the use cases Axiata has already deployed. It is now looking to deploy SMO by the end of this year, with the target of it being operational by the end of 2026. “It is a bit of a leap; we haven’t seen any stable SMO from vendors, but we believe in it because the emerging rApps ecosystem, which will replace traditional SONs [self-organizing networks] and will give us a major advantage on the network side,” Hundt said during a panel discussion.
Globally, there have been only limited SMO deployments.
What stands between Axiata and AI-native operations
The path to becoming an AI-native telco and building out autonomous networks is not without its challenges.
The company is taking a hard stance against vendors who maintain so-called data black boxes. “There are certain vendors that are still not open enough. For us, vendors who will not comply with the concept of making data truly available anytime without keeping it inside boxes, will not have a future at Axiata,” Hundt said.
In addition, building data and ensuring its availability in real time is a major challenge. “The second part is to truly build a unified data ocean that brings together all types of data — network data, customer data, telemetry, and more — that a telco enterprise generates. This must be orchestrated and stored intelligently, without escalating costs, ensuring that the right data is available and properly retained in real time or near real time. Only then can monetization opportunities emerge — for example, through hyper-personalization,” Hundt elaborated.
To achieve its vision of an AI-native telco, Axiata needs to find people with the required skillsets, which is another challenge it has to contend with.
“No people means no AI and no data,” Hundt said. “Building these competencies, grooming talent and fostering the right mindset remain the key challenges. It is very, very difficult to find the right people. We are operating in markets where the basic educational setup might not always be the most advanced one. In general, the availability of people is difficult. We need to invest a lot into postgraduates and then in developing specific telco technology and AI capabilities.”