The telco’s CTO, Desmond Cheung, told RCR Wireless News that the AI-RAN Research Center fundamentally shifts Indosat’s network strategy from being connectivity-led to AI-native
In sum – what to know:
Indosat shifts to an AI-native network strategy – The AI Grid links local AI hubs to its AI Factory, enabling real-time processing, lower latency, and smarter applications across Indonesia.
AI-for-RAN and AI-on-RAN drive early value – AI will automate traffic and energy management while enabling services like Sahabat AI, Vision AI, and edge applications for education, healthcare, and industry.
The AI-RAN Center accelerates digital sovereignty – The facility supports training, experimentation, and developer collaboration to build Indonesia’s homegrown AI and telecom talent base.
Indonesian operator Indosat Ooredoo Hutchison is steering its entire network strategy toward an AI-native future — a shift that, according to the company’s director and CTO Desmond Cheung, will redefine how intelligence is delivered across Indonesia.
In an interview with RCR Wireless News, Cheung said the company’s new AI-RAN Research Center is the engine behind this transformation, enabling Indosat, Nokia, and Nvidia to build what he calls a nationwide AI Grid that brings real-time AI processing directly into the network.
“The AI-RAN Research Center fundamentally shifts Indosat’s network strategy from being connectivity-led to AI-native,” Cheung said, explaining that the AI Grid will connect AI hubs across the country using Indosat’s RAN footprint and centralized AI Factory. With more than 56,000 sites, he added, Indosat is positioned to “activate an AI Grid quickly and efficiently.”
Cheung described the AI Grid as a system designed to run both “real-time AI processing closer to customers and heavier AI workloads in our central AI Data Center,” supporting use cases ranging from AI-for-RAN automation to Vision AI and, eventually, Physical AI for industrial sectors.
On early use cases, the CTO highlighted two parallel tracks: AI optimizing the network itself, and AI delivered through the network. “AI will automatically manage traffic, improve signal quality, and help us use energy more efficiently — giving users a smoother and more reliable connection wherever they are,” he said. At the same time, Indosat plans to run applications such as its locally trained Sahabat AI LLM directly on the RAN infrastructure, enabling education, healthcare, and enterprise services at the edge.
Cheung emphasized that the AI-RAN Research Center is as much a talent and innovation engine as a technology platform. “The AI-RAN Research Center is the heart of Indonesia’s innovation in intelligent connectivity — a place where ideas, technology, and talent come together,” he noted.
“Beyond research, the AI-RAN Center plays a vital role in strengthening Indonesia’s digital ecosystem. It connects global technology leaders such as Nokia and Nvidia with local universities, startups, and innovators — fostering collaboration, knowledge sharing, and the development of local expertise,” the executive said.
“Our lab provides a test environment that helps developers and startups reduce development costs while accelerating the creation of meaningful, high-impact applications. This lowers the barriers to innovation and supports the growth of Indonesia’s AI and digital solutions ecosystem,” he added.
Looking ahead, Indosat sees its phased deployment — beginning with four AI Grid clusters over the next months — as the foundation for a scalable national AI infrastructure. “Our focus is on identifying high-value use cases based on market needs and delivering AI-enabled, high-impact digital solutions for industries and customers,” Cheung said.
In July, Indosat Ooredoo Hutchison had partnered with Cisco and Nvidia to support the creation of Indonesia’s AI Center of Excellence (CoE).
