FEATURE
The broader implication is that data architecture can no longer be treated as a separate infrastructure concern. It is becoming a central determinant of AI performance, reliability and scalability.
Agentic AI changes the operating model
The emergence of agentic AI represents another turning point in enterprise adoption.
While generative AI applications typically respond to prompts, agentic systems are designed to pursue objectives, interact with applications and coordinate multi-step activities.
This creates opportunities to automate more complex business processes, but it also introduces new requirements around governance, security and accountability.
The broader implication is that data architecture can no longer be treated as a separate infrastructure concern. It is becoming a central determinant of AI performance, reliability and scalability.
He believes production AI agents should operate with clearly defined human ownership, approved workflows and appropriate supervision.
defined boundaries.
Globant ' s AI Pods model reflects this philosophy by combining agentoperated service units with human expertise.
Amjed describes the emerging enterprise model as controlled autonomy.
Under this approach, intelligent agents can retrieve governed information, interact with approved applications and execute tasks within
This requires secure APIs, granular access controls, orchestration platforms, observability and auditable decision trails. Human oversight remains particularly important for decisions involving financial, operational or regulatory consequences.
Al-Anqari believes the transition will fundamentally change how AI operates within enterprises.
" The shift toward agentic AI represents a change in how enterprises operate. We are moving from AI that responds to human prompts to AI that autonomously pursues objectives, orchestrates multi-step workflows, and makes consequential decisions in real time," he said.
Such capabilities require trusted data environments, secure integration and infrastructure capable of supporting reliable execution.
For Pienovi, agentic AI is particularly significant because it enables organisations to move beyond automating individual tasks towards coordinating entire processes.
Globant ' s Agentic Tourism Corridor demonstrates this approach by allowing intelligent agents representing different organisations to coordinate services across a connected ecosystem.
However, Pienovi emphasised that automation does not eliminate human accountability.
" When a task moves to an agent, the task leaves someone ' s day, but ownership of the outcome does not," he said.
Buzzalina expects employees to increasingly interact with AI as a digital teammate rather than a standalone application.
Instead of navigating multiple systems, workers could use intelligent agents to retrieve information, initiate activities, monitor processes and resolve operational issues.
Ayman believes the transition will also require organisations to reconsider how work is divided between people and machines.
Agents can manage routine execution and present relevant information, while employees concentrate on judgement, exceptions and situations requiring deeper contextual understanding.
DXC describes these moments of human intervention as " good friction ", recognising that not every decision should be automated.
For enterprises, the challenge will be determining where autonomy creates value and where human oversight remains essential.
Turning AI investment into measurable returns
Despite growing enthusiasm for AI, demonstrating return on investment remains a critical challenge.
38
INTELLIGENT CIO MIDDLE EAST www. intelligentcio. com