You’ve led AI and data transformations across radically different industries, from J&J’s $35 billion supply chain to Cisco’s GTM engine to G42’s sovereign AI platform. What’s your mental model for distinguishing which AI investments actually move the needle versus those that generate excitement but only limited enterprise value?
I tend to start with the business problem rather than the technology. The question is not “Where can we use AI?” but “Where can AI fundamentally change an economic outcome?”
I usually look at three things: the size of the business problem, the degree to which AI can create a step change rather than a marginal improvement, and whether the organization can actually operationalize the solution at scale.
The strongest AI opportunities typically sit at the intersection of material business value, proprietary data or context, and a workflow where AI can meaningfully change how work gets done. That could mean increasing conversion, reducing churn, improving underwriting, accelerating software development, or dramatically improving productivity.
I am also increasingly skeptical of AI use cases where the model is impressive but the workflow remains unchanged. A great model embedded in a weak process creates a demo. A good-enough model embedded in the right workflow can create a business.
Having built and led teams of 650–750+ engineers across Singapore, India, and the US, how has your approach to scaling global, matrixed data organizations evolved—and what’s the hardest leadership lesson you’ve had to learn the hard way?
Early in my career, I probably over-indexed on getting the organization, processes, and architecture exactly right. Over time, I have learned that the ability to attract and retain talent and operating with scale, clarity, and speed matter equally as organizational perfection.
I now spend much more time defining a small number of non-negotiable outcomes, creating clear ownership, and giving strong leaders room to operate. Global teams work particularly well when you optimize for complementary strengths rather than trying to make every location identical.
The hardest lesson was learning that leadership is not about being the smartest person in the room or having all the answers. It is about building an environment where the organization can consistently make good decisions without you.
As the organization grows, your job changes from solving problems to building the system that solves problems.
The strongest AI opportunities typically sit at the intersection of material business value, proprietary data or context, and a workflow where AI can meaningfully change how work gets done.”
Your decade at BCG gave you a front-row seat to how Fortune 100 executives think about transformation—and you’ve since sat on the other side of the table as a CDAO and VP across Cisco, J&J, and now Singtel. How has that BCG lens continued to shape the way you frame strategy, prioritize AI investments, and drive large-scale programs from the inside?
BCG fundamentally shaped how I think about problems.
The biggest lesson I took away is the importance of starting with the “so what?” Before getting into technology, data, or organizational design, I try to establish what outcome we are trying to change, how much it matters, and what would have to be true to achieve it. It also taught me to ask the hard questions, focus on the right problems, and simplify solutions so that they can scale rapidly.
That discipline has stayed with me throughout my operating career. The difference now is that I have to live with the consequences of the strategy. At BCG, you could recommend the right answer. As an operator, you also have to build the organization, secure the funding, deal with legacy technology, manage change, and deliver the result.
That experience has made me much more pragmatic. Strategy matters, but strategy without execution architecture is just a presentation.
Having worked at the frontier of agentic AI, sovereign platforms, and enterprise data transformation across multiple continents, where do you see AI heading over the next five years and which developments do you think are most underappreciated by business leaders today?
I think we are moving from AI that answers questions to AI that actually does work.
The next phase will be defined by agents operating across enterprise systems, reasoning over proprietary context, and increasingly executing multi-step workflows. The real competitive advantage will therefore shift from access to models to the quality of the surrounding ecosystem: data, context, memory, tools, workflows, governance, and the ability to deploy agents safely at scale.
I also think the economics of AI will change dramatically. As models become cheaper and more capable, the bottleneck will increasingly be organizational rather than technological.
One of the most underappreciated shifts is that AI will reshape the architecture of the enterprise itself. We spent decades building systems around applications and databases. We are now going to build much more of the enterprise around intelligence, context, and agents.
Before getting into technology, data, or organizational design, I try to establish what outcome we are trying to change, how much it matters, and what would have to be true to achieve it.”
As Singtel’s chief AI, data and analytics officer, you’re now leading AI transformation at one of Asia’s largest telcos. What are the most consequential opportunities and the thorniest challenges you’re focused on in the early months of that role, and how does the telco context differ from the enterprise tech environments you’ve navigated before?
Telecom is a fascinating AI environment because the combination of scale, customer interactions, network complexity, and real-time decisioning creates enormous opportunities.
I see the opportunity to power all customer interactions, network decisions, growth engines, and employee experience through agentic AI, and ultimately create entirely new AI-enabled products and business models.
One of the biggest opportunities is to make AI deeply embedded in day-to-day decision-making rather than treating it as a collection of standalone use cases. The ambition is to make intelligence a capability that can be reused across the enterprise.
The challenge is equally interesting. Telcos are large, complex businesses with significant legacy environments, huge volumes of data, and many interconnected systems. The hard part is not simply building a model or an agent. It is making AI reliable, secure, explainable, and scalable in an environment where the consequences of getting things wrong can be significant.
What makes Singtel particularly exciting is the opportunity to think beyond one company and build capabilities that can potentially create value across a broader telecom ecosystem in Asia.
Looking back across BCG, IIT Bombay, IIM Bangalore, and two decades of building data-driven enterprises, is there a piece of advice you wish someone had given you earlier in your career that you now find yourself passing on to the next generation of technologists and business leaders?
I would probably say, don’t optimize your career for the next title. Optimize it for the problems you get to solve and the people you get to work with.
Value is unlocked at the intersection of business and technology. Throughout my career, I’ve found that the people who create the most impact are those who can understand both sides and connect technology to a business outcome that matters. The most valuable technology leaders are not the ones who know the most about technology. They are the ones who can connect technology to an outcome that matters.
Technology will change, industries will change, and roles will change. The ability to learn quickly, simplify complex problems, communicate clearly, and build high-performing teams compounds much faster than any specific technical skill.
My advice to the next generation is to stay curious about technology and get your hands dirty. You don’t need to become an expert in every new technology, but you should understand what is changing, experiment with it yourself, and develop a point of view on where the world is going. I also try to invest around 10% of my time in learning what’s next. That could be a new technology, a new business model, or simply understanding how a market is evolving. That investment compounds over time and helps you stay ahead rather than react after the world has already changed.
And perhaps most importantly, choose the hard problems. The situations that look messy, ambiguous, and almost impossible are often the ones that teach you the most. Technology will continue to evolve faster than most organizations can adapt. The ability to keep learning, connect technology to business value, and turn that understanding into action is what I believe will differentiate the leaders of the next generation.
