Indian enterprises have made significant progress in digitization over the past decade, accelerated further by cloud, data, and AI adoption. However, true transformation goes far beyond layering technology onto existing processes. It requires a fundamental rethink of business models, operating structures, and leadership accountability. In this conversation, we explore the mindset shifts, sectoral inflection points, and execution disciplines required for enterprises in India to move from incremental digital adoption to meaningful, enterprise-wide transformation.
1. In your view, what are the biggest mindset shifts Indian enterprises still need to make to truly transform rather than just “digitize at the edges”?
Many Indian enterprises have digitized fast, but the leap from “digital” to “transformed” is a different game. It starts with moving from tech adoption to business reinvention—not just automating yesterday’s processes, but rethinking how value is created, how decisions are made, and how the organization competes.
A common blocker is the “pilot trap.” Teams run multiple proofs of concept, but they don’t hard-wire a path to scale. The shift is from proof of concept to proof of value: pick a handful of value streams, such as order-to-cash, claims, plant reliability, or customer onboarding, redesign them end to end, and scale what works with discipline.
Another mindset shift is moving beyond cost-only thinking. Digital and AI can absolutely take cost out, but leaders who achieve disproportionate outcomes use them to drive growth, resilience, and differentiation. That also changes ownership: the business must be accountable for outcomes, with the CEO and business leaders driving it as a business agenda, not treating it as “IT’s program.”
And none of it scales without getting the basics right: clean and connected data, clear data ownership, access controls, auditability, and security by design. If the foundation is weak, AI initiatives either slow down under risk and compliance questions or scale in ways that create operational and reputational exposure.
Net-net, the advantage comes from learning and adapting faster than the market. The enterprises that build feedback loops, collaborate across business and technology, and are willing to redesign the operating model—not just the apps—are the ones that move beyond digitizing at the edges.
2. India has a unique mix of large legacy enterprises and fast-growing digital-native firms; how does your transformation playbook differ when you work with each of these archetypes?
In India, the playbook needs to be flexible because “legacy” and “digital-native” firms start from very different constraints.
With large legacy enterprises, the work is usually core reinvention. You’re dealing with decades of process layers, multiple ERP and custom systems, and data spread across multiple business units. So, we start by simplifying the operating model, modernizing the data backbone, and aligning transformation to a few business outcomes that the CEO and business heads will sponsor. The hard part is often governance, compliance, and change—for example, decision rights, funding models, and adoption across plants, branches, or regions—not the technology itself.
Digital-native firms have different growing pains. Speed is their advantage, but as they scale, they run into reliability, security, unit economics, and regulatory expectations, especially in BFSI-adjacent or consumer-facing businesses. Here the focus shifts to resilient architecture, stronger governance, better FinOps, and production-grade controls, without slowing the product cadence that got them here.
The end state is similar in both cases: an AI-driven, outcome-led enterprise. The route is what changes. Legacy organizations need to simplify and accelerate the core. Digital natives need to industrialize and govern at scale. When we get that sequencing right, transformation stops being a collection of programs and becomes a capability the business can repeat.
3. With experience spanning telecom, manufacturing, automotive, healthcare and IT services, which sector do you believe is at an inflection point for large-scale transformation in India? What is NTT DATA doing differently there to move the needle?
Manufacturing is at one of the clearest inflection points for large-scale transformation in India right now. Supply-chain realignment, policy momentum including PLI, and global competitiveness pressures are pushing the sector beyond scattered Industry 4.0 projects toward enterprise-wide reinvention. AI is no longer an add-on; it’s starting to sit inside how plants run and how decisions get made across the value chain.
What feels different in this cycle is the move from isolated “smart factory” use cases to connected workflows across planning, production, quality, maintenance, and supply chain. For example, when demand planning and production scheduling are linked to quality signals and maintenance data, you see a step-change in throughput, inventory, and customer OTIF.
At NTT DATA, we focus on making transformation repeatable at scale. That means strong data and platform foundations, practical OT and IT convergence, and embedding AI into day-to-day decisions on the shop floor and across functions.
Execution discipline is where many programs win or stall. Leaders translate early wins into standards, reusable components, and a clear change plan so plants and teams adopt consistently. The goal is not “more digital.” It’s connected, intelligent operations that improve productivity, quality, resilience, and global competitiveness.
4. In recent times, the discussion is all about generative AI as a strategic revolution, not just a technology upgrade; what are the most compelling use cases you are seeing in India and APAC today, and where do you think leaders are still underestimating GenAI’s impact?
Across India and APAC, GenAI has moved past curiosity projects. The strongest results show up when it’s embedded into core workflows, with clear guardrails, rather than used as a standalone productivity assistant.
The use cases showing the most traction include customer service and experience, with multilingual agents for India’s language diversity, faster resolution, better agent assist, and lower cost-to-serve. Other strong areas are operations and enterprise productivity, software engineering and IT operations, and marketing, sales, and knowledge work. We are also seeing early agentic AI use cases where agents coordinate tasks across systems and move from assistance toward controlled execution.
Where leaders still underestimate GenAI is the operating-model change behind it. Copilots and efficiency gains matter, but the bigger lift is redesigning end-to-end processes, setting clear human review and escalation paths, and putting governance around data access, IP, and model risk.
Without that, GenAI stays trapped in pockets, or it scales in ways that make risk teams understandably nervous. So it’s not just a technology decision. It’s a leadership decision about how work runs, how risk is owned, and how value is measured.
5. Financial services and BFSI have been early adopters of AI-driven fraud detection, risk management and hyper-personalization; from your recent engagements, what patterns are emerging around what separates successful AI adopters from laggards?
In BFSI, simply “using AI” is no longer the differentiator. What separates leaders from laggards is disciplined execution, the ability to scale under regulatory scrutiny, and clarity on who owns risk.
Leaders also connect AI to end-to-end journeys, not isolated models. For instance, fraud signals, customer communication, and operations playbooks sit in one flow so action follows insight. That’s the real gap: redesigning workflows around measurable business impact.
Data readiness matters as much as model sophistication. The winners invest early in integration, real-time pipelines, and platforms built for auditability and access control. In regulated environments, weak foundations become a scaling constraint quickly, especially when explainability, logging, and jurisdictional controls are required.
Workflow redesign is the other big separator. High-performing institutions embed AI directly into decision flows for credit, fraud, compliance, and relationship management, rather than bolting it onto legacy processes. They are also more decisive. Speed helps, but only when it is paired with governance and clear accountability.
Responsible AI is not a side topic in BFSI; it’s how you earn the right to scale. Governance works best when it is designed to enable speed safely, not to create extra layers of approval. Finally, leaders treat AI as a business capability, not an IT project, and invest in adoption, role changes, and sustained change management.
6. Data readiness and responsible AI keep coming up in conversations about GenAI adoption; if you had to give CXOs a simple, three-point readiness checklist before embarking on large AI programs, what would be on that list?
If CXOs want GenAI to move beyond demos, the readiness question is simple: can you trust the data, control the risk, and run it like a production capability?
First, make the data foundation production-ready. Modernize your data architecture and make sure data is connected, governed, and usable across the business, not locked in functional silos. This means improving quality, being explicit about ownership, and applying automation to the unglamorous work too, such as classification, lineage, quality checks, and policy enforcement.
Second, tie use cases to outcomes and assign owners. Avoid a long list of “interesting ideas.” Pick a small set of high-value use cases, define what success looks like, and name the business owner who will carry it. The technology team enables, but the business must own adoption and outcomes.
Third, build the operating model, not just the model. Decide upfront how models will be deployed, monitored, and improved. Put in place MLOps or LLMOps, security controls, human review paths for sensitive decisions, and responsible AI governance that is proportionate to the risk. Then embed AI into the workflow itself so it drives outcomes rather than becoming another dashboard.
If those three are in place—data foundation, outcome ownership, and a production operating model—you can move from experimentation to scaled impact with far fewer surprises.
7. Sustainability, inclusivity and diversity are becoming core to how global enterprises define success; how do you see these themes intersecting with cloud, AI and infrastructure transformation in the Indian context over the next three to five years?
In India, sustainability, inclusivity, and diversity are increasingly part of the transformation brief, not a separate ESG track. As cloud and AI footprints grow, enterprises will run into real constraints and expectations around energy, water, skills, and equitable access.
On sustainability, infrastructure choices will be judged on efficiency as much as scale. As AI increases compute demand, enterprises will need greener architectures, for example workload placement, higher utilization through AIOps, cleaner power sourcing, and better cooling and water efficiency. Measurement matters here: if you can’t baseline and track emissions and energy per workload, it’s hard to manage it.
On inclusivity, cloud and AI can expand access at scale, but only if we design for India’s diversity. Multilingual experiences, assisted service models, and mobile-first journeys can bring more people into formal financial services, healthcare, and government services. The caution is important too: inclusive outcomes depend on responsible data practices and model design that reduce bias, plus strong privacy controls so trust is not compromised.
Diversity shows up in practical ways: who is building the systems, what data the models learn from, and how they are tested. Representative datasets, bias testing, and accountable Responsible AI practices are essential if AI systems are to reflect India’s social and economic reality rather than flatten it.
Put simply, the next phase of transformation will reward organizations that engineer for sustainability, design for inclusion, and treat trust and responsibility as non-negotiables.
8. Transformation is often discussed in terms of technology and operating models, but not enough in terms of leadership capabilities; what new leadership behaviors do you believe are non-negotiable for CXOs in the GenAI era?
In the GenAI era, the constraint on transformation is increasingly not technology, but leadership behavior. When AI ambition turns into sustained business value, it’s usually because leaders are visibly engaged, make decisions quickly, and stay accountable when trade-offs show up.
Personal ownership from the top is non-negotiable. AI must be treated as a core business priority led by the CEO and executive team, not delegated as an IT initiative.
Leaders also have to sponsor end-to-end reinvention, not tool rollouts. That means changing workflows, decision rights, incentives, and KPIs so teams can actually work in an AI-enabled way. If the operating model stays the same, AI becomes a sidecar.
Responsible AI leadership is equally critical. Trust, governance, and risk ownership have to sit at the top, with clear escalation paths when models behave unexpectedly.
Finally, CXOs need to invest in people with the same seriousness as the technology. Upskilling is part of it, but so is role redesign, adoption support, and creating confidence in human-AI collaboration. Transformation doesn’t happen because tools exist; it happens when people change how they work.
9. Many organizations struggle to move from pilots and PoCs to scaled transformations; what governance or operating model changes have you seen make the biggest difference in moving from experimentation to enterprise-wide impact?
When organizations move from pilots to enterprise-wide impact, it’s usually because they make a few practical governance and operating-model choices that remove friction and create reuse.
One is enterprise-level, business-led AI governance. The teams that scale well typically run an AI portfolio like a business portfolio: a small number of priorities, clear funding, agreed standards, and explicit risk ownership. In many cases that means executive-level accountability and board visibility for the highest-impact, highest-risk use cases.
Another shift is moving from “more use cases” to value-stream ownership. Instead of funding dozens of disconnected pilots, leaders pick a few domains where redesign will matter, such as onboarding, claims, procurement, plant reliability, or collections, then rebuild the workflow end to end. That focus is often what unlocks scale.
A third differentiator is running AI like a production capability. Shared data platforms, reusable components, MLOps or LLMOps, security controls, monitoring, and Responsible AI checks make it possible to deploy, observe, and improve solutions consistently. Without this, every pilot is effectively a one-off build.
Finally, adoption needs to be treated as change, not rollout. That means updating KPIs, training teams, redesigning roles where needed, and building trust so people use AI in daily execution rather than as a novelty tool.
The core lesson is straightforward: scaling AI is less about running more projects and more about building an enterprise capability that keeps getting better over time.
The path from digitization to true transformation is less about deploying more technology and more about rethinking how the enterprise operates, decides, and creates value. Across sectors, the organizations pulling ahead are those that combine strong data foundations, disciplined execution, and business-led ownership with a willingness to redesign workflows end to end.
As AI—and particularly GenAI—reshapes industries, the differentiator will not be access to technology, but the ability to scale it responsibly, embed it deeply into operations, and align leadership, governance, and talent around it. In that sense, transformation is no longer a one-time initiative; it is an organizational capability that must be continuously built, refined, and led from the top.
