Dr.Ramesh Shanmuganathan, Executive Vice President and Group CIO at John Keells Holdings, Director/CEO at John Keells IT and Non-Executive Director at Nations Trust Bank
After decades of transformation programmes — cloud migrations, ERP rollouts, AI pilots, automation initiatives, data platforms, cybersecurity overhauls — most enterprises have changed everything around the business while leaving the business itself fundamentally untouched.
They digitized the wrapper, accelerated the workflows, optimized the processes, and modernized the infrastructure. But they rarely reimagined how value itself should be conceived, orchestrated, delivered, and continuously evolved in an intelligent, real-time economy.
Dr Ramesh Shanmuganathan, Executive Vice President and Group CIO at John Keells Holdings, Director/CEO at John Keells IT and Non-Executive Director at Nations Trust Bank, says that era is now ending — not gradually, not incrementally, but structurally. What is emerging next is not another technology cycle, not another transformation programme, not merely the next phase of cloud or artificial intelligence. It is a reinvention of the enterprise itself.
He calls it “Service as a Software”, and he believes it will redefine how organizations are designed, how leadership operates, how services are delivered, and ultimately, how competitive advantage is created in the decade ahead.
Most transformation initiatives simply helped yesterday’s operating model move faster on newer rails. That era is now ending.
Could you describe what Service as a Software consists of?
Let me be precise about what Service as a Software is not. It is not Software-as-a-Service repackaged with a more fashionable label. It is not AI layered onto legacy workflows and marketed as intelligent operations. It is not another platform abstraction designed to optimize existing inefficiencies. Those approaches may improve efficiency at the margins — they may even improve customer experience incrementally. But they do not fundamentally reinvent the enterprise.
Service as a Software begins at a different threshold entirely. It begins when the service itself — not the system that supports it, not the process that delivers it, but the service — becomes software-defined, AI-orchestrated, continuously adaptive, context-aware, and capable of operating with autonomous intelligence at scale. Historically, software supported services. In this new era, software becomes a service. And when that inflection point is reached, the implications extend far beyond technology architecture.
The organization changes. Leadership changes. Governance changes. The very logic of hierarchy begins to weaken. This is not an incremental transformation. This is enterprise redefinition.
“The real question facing leadership teams today is not whether this future is arriving. It already has. The question is whether your organization is designing itself for it, or waiting for certainty that will never come.”
How is this wave of transformation different from the ones that came before?
Over the course of 35 years at the intersection of enterprise technology and organizational leadership, I have watched multiple waves of disruption arrive with extraordinary momentum. Client-server computing changed the operational scale. The internet transformed connectivity. ERP systems standardize enterprise workflows. Cloud redefined infrastructure economics. Automation accelerated execution. Each wave mattered. But every previous wave largely optimized the existing enterprise model rather than replacing it.
What is happening now is fundamentally different in nature — not simply in degree. Artificial intelligence introduces something enterprises have never previously possessed at scale: the ability for systems to reason, orchestrate, adapt, and execute outcomes dynamically without continuous human intervention. That changes the foundational assumptions upon which modern organizations were built. Traditional enterprises were designed around human cognitive limitations — departments existed because expertise needed concentration, and management layers existed. After all, information flow required coordination, and handoffs existed because no single individual could maintain enterprise-wide operational context simultaneously.
AI agents operate differently. They do not lose context. They do not require repeated instruction. They do not operate within functional silos unless organizations intentionally constrain them to do so. Connected intelligently, they can coordinate across functions, synthesize decisions, personalize engagement, optimize execution, and continuously adapt in real time. That is not automation. That is a fundamentally different operating model — and it is already beginning to reshape enterprise architecture globally.
AI agents do not merely automate tasks. They introduce reasoning, adaptation, and autonomous orchestration — and in doing so, they quietly dismantle the logic that built the modern organization.
Why do you say the future enterprise stack will be intelligence-centric?
One of the greatest strategic mistakes organizations are currently making is attempting to deploy intelligent systems on top of static architectural foundations built for a previous era. The industrial-age enterprise was optimized for predictability, standardization, and control. Those were legitimate priorities when the environment was predictable, when markets moved at a pace quarterly planning could manage, when competitive disruption arrived with some warning.
That environment no longer exists. Markets shift in real time. Customer expectations evolve continuously. Cyber threats mutate daily. Innovation cycles compress relentlessly. Competitive disruption increasingly emerges from outside traditional industry boundaries altogether. In this landscape, static architectures do not merely underperform; they expose the enterprise to structural vulnerability that no amount of operational efficiency can compensate for.
The future enterprise stack, therefore, cannot remain application-centric. It must become intelligence-centric — built not around fragmented systems managed independently, but around intelligent service ecosystems anchored in real-time data architectures, AI orchestration layers, autonomous decision engines, composable APIs, embedded cybersecurity intelligence, and context-aware interaction frameworks. In this model, the enterprise no longer behaves like a collection of disconnected systems. It behaves like a living, continuously adaptive digital organism — responsive, context-aware, self-adjusting, and capable of autonomous evolution.
The implications are already visible in practice. Healthcare systems are beginning to shift from reactive treatment models towards predictive intervention ecosystems that anticipate clinical deterioration before the crisis occurs. Financial services are evolving from static product delivery to continuously personalized intelligence platforms that dynamically restructure guidance based on individual behavioural signals. Supply chains are self-optimizing across geopolitical, environmental, and economic variables simultaneously, in real time. Retail is transitioning from transactional distribution into real-time experience engines that personalize every interaction at machine speed.
These are not distant possibilities. The technological foundations exist. What remains scarce, in organization after organization, is not capability. It is leadership courage.
Does the transformation challenge lie more in leadership rather than technology?
Technology evolves exponentially. The thinking of most leadership teams evolves incrementally. And the widening gap between those two realities is precisely where transformation initiatives stall, not because the technology failed, but because the leadership did not evolve quickly enough to redesign around what the technology made possible.
When I sit with my leadership team at John Keells Holdings and look across our portfolio — retail, transportation, hospitality, financial services, property — the question I keep returning to is not which tasks AI can handle. The question is which of our services can become software-defined ecosystems, and what does our organization look like when they do. That reorientation of the question — from task automation to enterprise redesign — is the distinction that separates leaders who will define the next decade from those who will spend it catching up.
Hierarchy was never accidental. It emerged as a rational coordination mechanism in a world constrained by human cognitive limitations. But as AI agents dramatically reduce the friction that hierarchy was created to solve — when information flows freely, when coordination can be automated, when execution can be orchestrated dynamically across functions — the justification for excessive organizational layers weakens structurally.
This does not mean organizations suddenly become flat. It means leadership roles evolve fundamentally. Managers increasingly transition from process supervisors into strategic orchestrators, from information gatekeepers into decision enablers, from operational coordinators into culture architects. The leaders who recognize this shift early and begin developing that next version of themselves will remain highly relevant. Those who continue defending industrial-era management models inside intelligent enterprises will find it progressively difficult to justify the organizational complexity that technology has already rendered unnecessary.
Leadership readiness matters more than technology readiness. The technology is arriving regardless. The question is whether leadership evolves quickly enough to redesign around it.
Why is trust so important in this transformation, and how do you define it within the intelligence era?
As enterprises become increasingly software-defined and AI-driven, one principle becomes absolutely non-negotiable — and it is not the one most organizations prioritize first. It is trust. Not trust as a brand value. Not trust as a marketing message. Not trust as a compliance checkbox reviewed annually and filed accordingly. Trust as engineered architecture, designed in from the beginning, not bolted on as an afterthought.
When intelligent systems begin influencing healthcare outcomes, financial decisions, customer access, and operational governance, trust ceases to be a communications problem. It becomes a systems-design problem. Every API in a software-defined enterprise is a trust boundary. Every AI model is a governance challenge. Every autonomous workflow is a potential risk vector. Every customer interaction is a data responsibility.
Cybersecurity in this context cannot remain a perimeter defence function. It must become embedded operational intelligence woven directly into enterprise architecture itself — present at every layer, not applied after the fact.
The organizations that succeed in this next era will not merely deploy intelligence effectively. They will combine intelligence, transparency, resilience, governance, and trust simultaneously, treating them not as sequential priorities but as simultaneous architectural commitments. Because consumers will not tolerate opaque systems indefinitely. Regulators will not tolerate ungoverned AI. Boards will not tolerate uncontrolled operational exposure. And markets, ultimately, do not reward fragility.
The future winners will not be the organizations with the most advanced technology. They will be the organizations capable of combining intelligence with trust — and making that combination their defining competitive differentiator.
Where do emerging economies stand in this environment?
There is a dimension to this transformation that deserves far greater attention than it currently receives, particularly for emerging economies like Sri Lanka. Historically, developing markets operated at structural disadvantage: limited infrastructure, capital constraints, technology lag, and dependency on imported operational models. The playbook of global enterprises was always several years ahead of what local organizations could realistically execute.
That asymmetry is now collapsing — and in ways that may favour those who have historically been disadvantaged. Organizations burdened by decades of deeply entrenched legacy systems are not, in this new environment, necessarily the most advantaged. In many cases, they are the most constrained. Sunk costs in outdated architecture. Cultural resistance embedded in decades of operating model inertia. The weight of what worked yesterday is slowing the adaptation to what is needed tomorrow. Technical debt, in the era of Service as a Software, has become strategic debt.
This creates a rare and time-limited opportunity. Organizations with fewer entrenched legacy systems — and the leadership courage to move decisively — can leapfrog directly into cloud-native, AI-first, composable operating models. The absence of technical debt, reframed, is a competitive asset. This is Sri Lanka’s window. Not a distant aspiration. A present opportunity that requires present-tense decisions.
The next global competitive landscape will not be determined purely by industrial scale or historical market position. It will be shaped by digital adaptability, AI readiness, cybersecurity maturity, innovation velocity, and above all, the quality of leadership vision.
We have the talent. We have the ambition. What the moment demands is the courage to stop benchmarking against our past and start designing for the future that is already arriving.
What will the enterprise of the future look like?
Over the next decade, the distinction between technology companies and non-technology companies will not blur. It will disappear. Banks will increasingly behave like intelligent platforms. Healthcare providers will become predictive service ecosystems. Retailers will evolve into real-time experience engines. Manufacturers will operate through autonomous operational networks. Hospitality will become context-aware service intelligence. Every industry will function, to an increasing degree, through software-defined service layers powered by intelligent agents, orchestrated by real-time data, and governed by trust.
This is the new enterprise stack — not infrastructure alone, not applications alone, not AI alone, but intelligent, orchestrated, continuously adaptive service ecosystems designed around outcomes, built on trust, and capable of real-time evolution. In that future, competitive advantage will not belong to organizations with the largest technology budgets or the longest transformation roadmaps. It will belong to organizations with the clearest leadership vision — leaders willing to redesign their enterprises around intelligence rather than efficiency alone, courageous enough to rethink service itself as intelligent software while preserving the deeply human foundations of trust, ethics, judgement, and purpose that every sustainable enterprise ultimately rests upon.
I have spent 35 years watching technology promise transformation and watching organizations choose optimization instead. What is arriving now does not offer that choice. The forces reshaping enterprise architecture — AI orchestration, autonomous service delivery, intelligent agents, real-time data ecosystems — are not features to be adopted selectively. They are environmental conditions to be navigated strategically. And the organizations that treat them as the former will find themselves disrupted by the organizations that understand them as the latter.
The real question facing leadership teams today is not whether this future is arriving. It already has. The question is whether your organization is designing itself for it or waiting for certainty that will never come.
The organizations that will lead the next decade are not waiting cautiously on the sidelines. They are already redesigning themselves — embedding intelligence into every service, every interaction, every decision, and every layer of the enterprise. Because in the era of Service as a Software, the organizations that move with clarity and conviction will not simply gain a competitive advantage. They will redefine the rules of the industry itself.