Unlocking ‘dark data’ crucial to agentic AI shift, says expert

Accessing the unstructured data scattered within corporate systems is the main obstacle for businesses transitioning to autonomous artificial intelligence (AI), according to an industry specialist. Ghassan Kosta, regional general manager for Qatar, Oman, Bahrain, and Iraq at Google Cloud, highlighted during an exclusive interview that autonomous AI agents depend solely on the quality and context of the data that powers them.

He explained that raw data without context is merely noise in enterprise systems, which is why Google Cloud has developed specialized architectures to convert isolated files into structured corporate knowledge. Much of the operational intelligence within organizations lies dormant in unstructured assets such as contracts, internal communications, and disorganized documents.

Kosta outlined the process of linking structured tables with unstructured assets to enable AI agents to fully grasp an organization's operations and landscape. To meet the substantial computational demands of the agentic AI era, Kosta disclosed that Google Cloud is increasing its local infrastructure capacity, three years after setting up the Google Cloud Doha region.

He emphasized that transitioning to autonomous workflows necessitates significant changes in technical abilities, requiring deliberate enhancements in compute, storage, and software resources within local data centers. In addition to expanding infrastructure, Kosta stressed the importance of incorporating stringent security protocols and governance frameworks from the early stages to protect enterprise workloads as systems grow.

"Ensuring the security of customers' workloads begins with robust AI governance, trusted guardrails, and control from the outset," he stated, adding that security measures should cover the entire operational lifecycle from development to execution through autonomous threat intelligence. These structural improvements align with wider public sector efforts to bolster national digital capabilities under the Digital Agenda 2030.

Kosta noted that public sector institutions are establishing dedicated testing hubs, including an innovation laboratory (iLab) in collaboration with the Ministry of Communications and Information Technology (MoCI), to enable developers and researchers to prototype multimodal applications before widespread deployment. The Civil Service and Government Development Bureau, along with the National Planning Council, are also utilizing graph technology and analytics infrastructure to model future workforce scenarios and enhance executive reporting.

Kosta emphasized that deploying advanced technology must be accompanied by comprehensive workforce training to ensure digital transformation translates into tangible business value. Kosta recalled that Google Cloud’s local strategy in Qatar began with the establishment of its Centre of Excellence in 2021, followed by the opening of its official cloud region in 2023.

To meet future workforce needs, Kosta highlighted the launch of the Google Cloud National Skilling Programme, a partnership with the Qatar Digital Academy (QDA) aimed at providing over 50,000 new learning opportunities across Qatar by 2030. Kosta concluded by reiterating that cultivating the necessary technical skills in local talent pools is crucial for fostering continuous innovation, operational creativity, and long-term project viability across regional enterprises.