Datadobi’s Matthias Nijs explains why data visibility, continuous governance, and control across hybrid environments are essential for enterprise AI readiness.
Enterprises are racing to embed artificial intelligence into everyday operations, shifting the challenge from managing the sheer volume of data to understanding where information resides and how it is distributed across increasingly complex IT environments. The change is particularly visible in the UAE, where organisations have rapidly adopted hybrid and multi-cloud architectures.
According to Matthias Nijs, VP of EMEA Sales at Datadobi, data fragmentation is becoming a more pressing enterprise concern than data volume itself. “A few years ago, volume was the headline concern; now it’s fragmentation,” Nijs said.
Data fragmentation challenges
Enterprise data is increasingly distributed across on-premises infrastructure, private clouds and multiple public-cloud providers, each with its own tools, permissions and potential blind spots. Fragmentation can make it difficult for organisations preparing their data for AI to establish what information they hold, where it resides, who owns it and whether it is suitable for use.
“In my experience across the region, it’s knowing where the unstructured data actually lives,” Nijs said. The problem is particularly pronounced with unstructured data, including file shares, scanned documents, video and sensor-generated information. Unlike structured information held in databases, unstructured data can be dispersed across sprawling hybrid environments, with limited metadata explaining what it contains or why it is being retained.
Organisations may be able to make decisions about individual datasets, Nijs said, but often lack visibility across the entire estate. “You can make a decision about a particular dataset, but if you don’t have visibility across the estate, you don’t necessarily know what you’re missing,” he said.
Greater visibility is becoming increasingly important as companies look to use existing enterprise data for AI applications.
For Nijs, becoming “AI-ready” starts well before an organisation deploys an AI model. The first step involves basic data hygiene: understanding what data exists, where it is stored, who is responsible for it and whether it is sensitive or subject to regulatory requirements.
“AI-readiness starts with data discipline, not with the model itself,” he said.
Organisations that overlook this foundation can encounter problems when AI systems interact with outdated, poorly assessed or inappropriate information, he added.
“AI is only going to be as useful as the data you allow it to access,” Nijs said. A model may surface information it should not be permitted to access or produce weaker results because the underlying data has not been properly assessed.
Continuous AI governance
The transition from AI experimentation to routine enterprise use will also reshape the responsibilities of CIOs and data-management teams, according to Nijs. “I think the CIO’s role becomes less about approving individual AI projects and more about governing the data estate those projects all draw from,” he said.
Data-management teams will need to move away from a reactive approach in which information is cleaned or organised only when required for a specific project. “Data governance can’t be something you do only when an AI project comes along,” Nijs said. “It has to become continuous.”
Continuous governance requires ongoing mapping, classification and oversight across the wider data estate. The UAE provides a particularly relevant environment for this shift. Nijs said the country has demonstrated an ability to turn technology strategies into infrastructure and adoption at speed, positioning it as a potential testing ground for large-scale AI and cloud initiatives.
“What stands out to me about the UAE is the speed at which strategy turns into infrastructure,” Nijs said.
“National AI and cloud ambitions here don’t stay on paper for long; they show up in real investment, real regulation and real adoption quickly.” Nijs expects the UAE’s position in the global technology landscape to evolve from that of a fast adopter into a market where large-scale AI and cloud strategies can be tested and developed before being adopted elsewhere.
Distributed infrastructure and AI’s growing dependence on enterprise data are making visibility across IT environments essential. Organisations need this insight to make informed decisions about which data should be retained, protected or made available to AI systems.
“The question is no longer simply how much data you have,” Nijs said. “It’s whether you understand that data well enough to trust it, protect it and use it.” For enterprises, the AI challenge may therefore be less about finding more data and more about establishing control over the information they already hold. “Before you ask what AI can do with your data, you need to know what data you actually have,” Nijs said.
Source: Tahawul Tech
