Beyond the Silos: The New Data Playbook for Accelerating AI in Indonesia

It’s a scenario all data leaders are familiar with: your CEO returns from an AI conference, buzzing about AI-enabled solutions. She or he wants chatbots handling customer service, predictive analytics optimizing inventory, and real-time insights flowing to every department — and wants it now. There’s just one problem: your company’s data is stored in 47 different locations, none of which communicate with each other.

This is a major reason why Indonesian enterprises are facing the so-called AI Stall.

Indonesia has set an ambitious target: it expects AI to contribute USD366 billion to the nation’s GDP by 2030. But bubbling underneath this optimism is a major hurdle. According to recent industry data, 81% of Indonesian companies cite poor data quality and availability as the single biggest barrier to AI implementation.

If Indonesia needs to move forward, it needs to address the AI Stall squarely. And it often begins with data.

Too many cooks, too many kitchens

“Indonesian businesses face data fragmentation,” explains an IBM Spokesperson in Indonesia. “Companies can’t access and analyze data wherever it resides without needing to move it.” The result? Projects stall, budgets drain and innovation stops.

It’s not that Indonesia lacks data. The nation sits on a mountain of valuable data. These come from customer transactions, supply chain records and social media interactions, right down to sensor readings from manufacturing floors. However, this data is fractured, siloed across departments and legacy systems that were never designed to share information. It’s like having all the ingredients for a feast scattered across a dozen locked kitchens.

Traditional approaches to fixing this mess involve massive data migration projects, consolidating everything into expensive centralized data warehouses. These initiatives can cost tens of millions of dollars and take years to complete. More importantly, they can still leave companies with rigid systems that aren’t agile enough for the next wave of innovation.

Keeping it local while going global

The Indonesian government clearly recognizes the stakes. It has actively invited global tech leaders to strengthen the digital ecosystem. This includes the development of the National Data Center (PDN), which aims to build AI capabilities without compromising data sovereignty.

“Their technology allows government and enterprises to keep sensitive data within national borders while still benefiting from cloud scalability,” notes the IBM Spokesperson about hybrid solutions being deployed. “This ensures compliance with local regulations and provides a smooth path from legacy systems to modern infrastructure.”

When your AI can’t read the room

But there’s a deeper problem than just where data lives. It’s about what kind of data modern AI actually needs. The next generation of AI applications (those that will deliver USD366 billion in value) do not just crunch numbers in spreadsheets. They need to understand documents, images, customer service chat logs, video content, and voice recordings. This “unstructured data” makes up roughly 80% of enterprise information, yet most Indonesian companies have no systematic way to prepare it for AI use.

IBM recognizes this hurdle. It offers IBM watsonx.data, which is built on an open lakehouse architecture optimized for governing data and AI workloads, with capabilities including querying, governance, and support for open data formats, supporting multiple query engines like Presto, Spark, Db2, and Netezza on a single governed platform. More recently, the company integrated DataStax Astra DB, a NoSQL and vector database built on Apache Cassandra, directly into the watsonx stack to manage unstructured data.

“By integrating DataStax Astra DB into watsonx, businesses can manage and search unstructured data — like documents, images, and chat logs — in real time,” the IBM Spokesperson explains. “This is essential for building GenAI applications that are accurate and context-aware.” Simply put, you can’t build a smart customer service bot if it can’t quickly find and understand the context buried in thousands of past support tickets.

The economics of IBM’s moves are also compelling. “Unlike traditional data warehouses that require large upfront investments, watsonx.data uses a flexible, pay-as-you-go model,” says the IBM Spokesperson. “It lets businesses choose the right tools for each job, which can cut data management costs by up to 50%.” In other words, it matches specific database engines to particular tasks, rather than forcing everything through a single, expensive system.

Humans are the other half of the answer

Of course, technology is only half the battle; the other half is human. Indonesian companies face a severe shortage of AI-specialized skills and governance expertise. This is where local system integrators and partner ecosystems become critical.

“IBM Partner Plus and local system integrators bring the expertise and training needed to help Indonesian companies implement AI responsibly,” notes the IBM Spokesperson. “They provide hands-on support, industry-specific solutions, and governance frameworks—bridging the talent gap.”

Perhaps more importantly, there’s the governance question. Indonesia’s regulatory environment around data is still evolving, creating uncertainty for enterprises trying to navigate compliance. “IBM watsonx includes built-in tools for data governance, helping businesses enforce policies and track data usage across all environments, cloud or on-premises,” the IBM Spokesperson explains. “With watsonx, companies can stay compliant while scaling their AI initiatives.”

Evolution, not revolution

Forget about ripping out existing systems and starting fresh. IBM sees the future in building bridges by connecting data where it sits, preparing it for AI use, and doing so in ways that respect regulatory boundaries while maintaining cost efficiency.

“With watsonx.data, companies can improve customer experiences, streamline operations, and make smarter decisions,” says the IBM Spokesperson, pointing to retail and telecom applications already showing results across the region.

Indonesia’s multi-billion-dollar AI ambition doesn't have to remain a dream deferred. However, achieving this requires confronting the data quality crisis head-on with hybrid architectures, proper governance, and the right combination of technology and expertise. The alternative is watching that massive economic opportunity slip away, one siloed database at a time.

Image credit: iStockphoto/Feodora Chiosea