Indosat is using AI to overhaul its operations – here's how

  • The operator is using a three-pillar strategy to use AI to transform its operations
  • IOH has formed several key partnerships, including with NVIDIA, Accenture and Nokia, to realize its AI vision
  • While IOH launched 5G in 2021, it continues to focus on 4G to address the needs of its subscribe

It seems like big bets on AI are paying off. Indosat Ooredoo Hutchison (IOH), Indonesia’s second-largest service provider, revealed AI helped it to bring down its capital expenditure by 22% to IDR9900 billion ($0.58 bn) last year even as it upgraded its network.

How, exactly? In an exclusive interview, IOH CEO Vikram Sinha said it's all about using AI to better allocate network resources. The service provider is leveraging AI to improve network operations, forecasting data traffic spikes, optimizing network capacity and quickly resolving network-related issues. 

“With AI, capacity planning accuracy can be 98%, and this enabled us to save around $10 million in CAPEX by doing capacity planning with AI," Sinha said. That same planning capability also helped boost the user experience by ensuring "excellent and continuous connectivity even during busy periods like we experienced Ramadan and Eid," he added.

Three-pillar AI strategy

IOH was formed with a merger between CK Hutchison and Ooredoo in 2022. It competes with Telkomsel and XLSmart Telecom, which was formed recently as a result of a merger between XL Axiata and Smartfren. 

The operator's overarching AI strategy is based around three pillars: become an AI Native Telco, operate as an AI TechCo and lead as an AI Nation Shaper. Its capacity planning effort was part of the AI Native Telco pillar, through which is it aiming to further transform operations via cognitive network management, IT optimization and improving customer experiences. 

And as for what it's doing on the other two fronts?

“One of the most important milestones was the establishment of Indonesia’s first full-stack sovereign AI factory, which we developed in collaboration with leading global partners including NVIDIA, Accenture, and Google,” Sinha said while explaining the AI TechCo pillar. This AI factory allows Indonesian enterprises to build, deploy, and scale AI applications within the country.

This was enabled by IOH’s subsidiary, Lintasarta, which became NVIDIA’s Cloud Provider Partner last year and launched GPU Merdeka, a sovereign AI-cloud platform providing on-demand access to NVIDIA’s GPUs and AI tools.

Shaping the AI ecosystem

About two years back, Indonesia came up with Digital Vision 2045, which outlines how AI will help the country advance its digital economy and governance.

As part of the AI Nation Shaper pillar, IOH is investing in Indonesia’s broader AI ecosystem by developing local talent and supporting the National Center of Excellence. A critical component of this part is building the Bahasa-based open-source Large Language Model (LLMs), Sahabat AI, in partnership with NVIDIA and GoTo Group. The model is trained on local Indonesian datasets.

“This means that Indonesians, whether students, researchers, SMEs, and businesses, can use it to build their own AI-based services that use the local language and are tailored for them,” Sinha explained. 

IOH has also partnered with Accenture to use its AI Refinery platform to help businesses build their own AI applications. “Initially, we are focusing on AI solutions for Indonesia’s financial services sector, one of the key pillars of the country’s economy, but we believe this is just the start,” said Sinha.

Launching AI-RAN

IOH was in the news recently for launching AI-RAN in partnership with Nokia and NVIDIA. It claims to be the first service provider in Southeast Asia and the third globally to deploy commercial AI-RAN infrastructure.

It combines Nokia’s 5G Cloud RAN with the NVIDIA AI Aerial platform, creating computing infrastructure that supports both AI and RAN workloads. 

“AI-RAN enables Indosat to significantly enhance network performance, energy efficiency and spectrum utilization by bringing AI capabilities directly into the radio access network,” Sinha said.

“It also aligns with Indonesia’s national AI strategy, enabling Indosat to serve AI inferencing needs at scale and unlock new revenue opportunities through AI-driven services in sectors such as healthcare, education and agriculture,” he added.

Interestingly, Asian service providers, including SK Telecom, Telstra, IOH and Singtel, among others, are betting on AI to go beyond offering voice and data services. They are emerging as frontrunners in the adoption of new technologies. However, the telcos face several challenges, such as attracting the right talent as well as evolving their operations and leveraging emerging technologies to transform themselves.

Focus on 4G and rural expansion

IOH launched 5G in eight cities, including Jakarta, Makassar and Bali, in 2021. The company is working with partners to develop new use cases and applications for 5G monetization. Even so, the service provider remains zeroed in on 4G because of limited 5G-device penetration.

“In Indonesia, 4G device penetration remains significantly higher than 5G. Therefore, our primary focus is on expanding and fortifying our 4G network, ensuring our stakeholders can continue to rely on our services,” said Sinha. 

To top it all off, IOH is also focused on growing its presence in rural areas. 

“In 2024, we invested around $800 million to strengthen our network, with around 60%, dedicated to fortifying our network in rural parts of Indonesia. In 2025, we are continuing to strategically invest in our network, particularly in rural parts of eastern Indonesia where we are currently under indexed,” Sinha concluded.