- With Phase 1 of its multi-cloud initiative, Cloud Infinity, complete, StarHub is now focused on monetization
- It is also facing challenges related to 5G monetization
- StarHub has started preparing for the impact of agentic AI on its workforce
Singapore’s second-largest service provider, StarHub, has spent the last two years on a cloud transformation journey. With the first phase of its overhaul now behind it, StarHub CTO Ayush Sharma told Fierce in an exclusive interview it is now moving towards the “harvest,” or monetization, phase.
StarHub launched its Cloud Infinity program in 2023 to transition from legacy systems to hybrid cloud-based systems. The operator used Red Hat OpenShift along with several cloud environments to achieve this.
“We wanted to use cloud in a differentiating way and that was essentially to build a cloud federation, run our infrastructure on that, but at the same time, partner with them and leverage the services of multiple cloud environments and create that offering with embedded connectivity and security to the customers,” Sharma explained.
Cloud Infinity builds on StarHub’s Dare+ strategy for cost rationalization and to transform the company from a telco to “a company that connects digital lives for customers.”
Already, Sharma said around 70% of StarHub’s control plane network runs in its new hybrid cloud. “30% we cannot move to Cloud Infinity because of regulatory constraints,” he added. “In addition, 90% of our content and 70% to 80% of our cyber products and services also run on this.”
With the first phase (dubbed “Build and Invest” by CEO Nikhil Eapen) of Cloud Infinity complete, the focus of the next phase is on “business model monetization and operating model," Sharma said.
“With these new platforms built, we now move to the ‘Harvest’ phase, benefitting from greater agility, scalability, and first-of-its-kind observability for long-term operational resilience,” Eapen said in a press release. “We are also able to drive scalable growth with cost optimisation. This is critical as we grow market share for our consumer business and for us to scale our Enterprise business regionally.”
Beyond using AI to optimize network performance and address tickets quickly, Sharma believes that StarHub’s cloud transformation is helping it leverage data that resides in its network to improve the customer experience.
“As a result of our transformation, we are able to go beyond hyper-personalization of the customers to create that virtuous cycle of as-a-service concept, which leads to leveraging the data in a more adaptive connectivity manner. This, in turn, programs the whole infrastructure based on the connectivity needs of the customer,” Sharma explained.
5G monetization and agentic AI
A technologically mature market, Singapore had around 9.9 million mobile subscribers (including 3G, 4G and 5G) and 12.7 million broadband subscribers, as of December 2024. The country has a vibrant telecommunications market with three major players —StarHub, Singtel and M1 — and comparatively smaller players, such as MyRepublic and ViewQwest.
One of the key challenges faced by StarHub is 5G monetization in this hyper-competitive environment.
“The industry is already hitting the maturity curve and 5G monetization continues to remain a challenge,” Sharma said. “If it’s a pure play connectivity without value creation, those products and services obviously are status quo. This is not a high-growth area for any telco and that includes us,” Sharma said. However, he believes StarHub’s federated cloud model allows it to evolve as a digital ecosystem enabler.
Additionally, Sharma is gung-ho about the innovations based on agentic AI.
“I am excited about agentic AI within the telco space, not for network-related, but about how we can create precise locations for the machines. I am excited about bringing autonomous adaptive services to these intelligent machines, be it humanoids, robots, or AGVs (automated guided vehicles),” he said, adding the operator hopes to showcase some of these innovations next year.
But at the same time Sharma acknowledged that agentic AI-led automation may lead to widespread workforce- and workflow-related changes across the organization since “many” roles may no longer be required. How the operator adapts to the changes AI will bring is “both a challenge and an opportunity to me,” Sharma said.
The company is trying to prepare its workforce, though, through a three-dimension training approach that spans experience, education and exposure.
“We have already trained over 300 engineers who are not only AWS or Google Cloud certified but are also involved in a variety of exposure through pilots and R&D initiatives, working with several companies like ServiceNow, OpenAI and NVIDIA, among others. This is critical to the success of not only building the top tech stack, but also tweaking the business and operating model for us,” he explained.
As far as the network is concerned, AI is also changing traffic patterns. Sharma called AI traffic the “next tsunami that is going to come,” but said StarHub’s network-as-software approach leaves it in “a much better position” to ride the wave and “auto-scale our network on steroids.”
Still, the rising tide of consumer and OTT traffic “will remain a challenge that will fill pipes pretty quickly, and as an industry, we have to figure out how to monetize that and then work with the ecosystem to drive the value for that massive data,” he concluded.