- IOH claims AI-driven automation has helped it to significantly reduce network downtime and customer complaints
- The Indonesian operator is expanding beyond telecom connectivity through AI factory infrastructure, GPU-as-a-service and sovereign AI initiatives
- IOH believes AI inferencing will drive the next major AI infrastructure wave
Indonesia’s second-largest service provider, Indosat Ooredoo Hutchison (IOH), is positioning itself as a regional AI infrastructure and neocloud provider, combining GPU compute, AI-native networks and sovereign AI services to capture the next wave of AI demand.
"We see this [GPUaaS and its Sovereign AI model, Sahabat AI] as an opportunity to evolve into a regional AI cloud, a 'neocloud', delivering sovereign and scalable AI capabilities not just for Indonesia, but across the region," said Chirag Sukhadia, chief data and AI officer at IOH. IOH was formed with a merger between CK Hutchison and Ooredoo in 2022 and competes with Telkomsel and XLSmart Telecom.
Its partnership with AI giant Nvidia is going to play a crucial role in its evolution as a regional player, acting as a "key enabler in this journey, giving us access to advanced GPU technology," says Sukhadia.
In addition, the country is “well positioned to evolve into a neocloud provider for Asia” because of its “strong fundamentals across land, power, and infrastructure," he said, adding, "The country offers abundant energy resources, including a growing renewable energy base, alongside large-scale land availability, making data centre expansion both scalable and cost-effective.”
What is a neocloud provider?
Neocloud providers exclusively focus on providing high-performance GPU-as-a-service (GPUaaS) and infrastructure for training and running AI models.
The market opportunity underpinning this ambition is substantial. Indonesia's AI-optimized data center market reached an estimated $660 million in 2025 and is forecast to nearly double to $1.44 billion by 2030, driven in part by the country's Personal Data Protection Law, which compels global providers to process AI workloads within Indonesian borders. This creates structural demand for sovereign compute infrastructure that IOH is building.
The full-stack vision combines GPU infrastructure with Sahabat-AI, an open-source large language model (LLM) optimized for the Indonesian language and local context, now available as a mobile chat service. The company is also building an AI Center of Excellence in partnership with tech companies Cisco and Nvidia.
Global cloud giants, such as Amazon Web Services (AWS), Microsoft and Google, already dominate the AI ecosystem through massive GPU procurement scale, mature developer platforms, enterprise relationships and vertically integrated AI stacks. Telcos, like IOH, are likely to face several challenges, including high capital investments and ensuring high utilization rates to remain economically viable. On the other hand, telcos come with significant advantages, including sovereign AI positioning, nationwide connectivity and distributed edge infrastructure.
Several other telcos in the APAC, including Singapore’s Singtel, are also targeting the region’s AI infrastructure market to grow their revenue and position in the AI ecosystem.
Apart from IOH’s neocloud play, the company has undertaken AI-native transformation of the core telecom business that is already producing significant results, including around 50% reduction in network downtime and customer complaints.
Now, the company is deploying an end-to-end agentic framework across its entire network operations to ensure “faster recovery and more reliable service for customers.”
“We are now enabling the end-to-end Agentic framework on the entire network operations, which will help us detect issues, initiate response and track resolution in real time,” said Sukhadia.
IOH was one of the first operators to announce its intent to become an AI-native telco in 2024. Its transformation journey is focused on three broad pillars of business value-led domain roadmap, data and tech platform and talent and adoption.
“The Data and Tech Platform is at an advanced stage, with all the relevant reusable capabilities already developed on the AI platform. Our people-centric approach with training our talent and driving the adoption across the organization is well-oiled now and is in a scale-up stage," said Sukhadia. "From a Value and Domain perspective, our Hyper Personalization (Commercial) and Smart Capex (Network) are at advance stage, while other domains like Hyper Localization (Distribution), B2B and Customer Service are now being scaled up."
Consumer AI: personalization driving the numbers
The commercial rationale for the transformation is evidenced in Indosat's financials. The company reported a revenue of IDR15,200 billion ($0.87 billion) in the first quarter of 2026 — the company’s highest quarterly revenue in its history. It recorded a profit of IDR1.5 trillion ($0.08 billion), growing 26% year-on-year. A key part of this growth is the AI-driven hyper-personalization, which allows IOH to offer the right product, at the right time, through the right channel, to its subscribers.
On the infrastructure side, Sukhadia believes that though the current demand is heavily driven by training workloads, the next phase will come from inferencing, as AI applications move into production at scale, closer to end users to ensure low-latency. This is in line with the global trend.
Indosat's answer is its AI Grid, an architecture that distributes AI capacity across the network rather than concentrating it in centralized data centers, reducing latency for inference applications and improving energy efficiency by placing compute closer to where demand originates.
For all the scale of Indosat's technical ambitions, Sukhadia believes the hardest part is not the technology but talent.
"The most challenging part of the transformation is not technology, but people and operating model change," he says. "Becoming AI-native requires rethinking how decisions are made, how teams work, and how we build capabilities at scale. It is about embedding AI at the design level for every process across all the domains, adopted by all employees."
The challenge is acute in the Indonesian context. Indonesia has set a target of training 90 million skilled tech professionals by 2035, and IOH has committed, through its Cisco and Nvidia partnerships, to training one million Indonesians in AI skills by 2027.
To learn more about how enterprises are extending their existing infrastructure to leverage AI, download our free report: "AI doesn't replace. It extends: A practical guide to retooling enterprise infrastructure and teams for the AI age."