The Digital Banker spoke with Rohit Puri, Senior Executive and Global Head of Business Process Services at NTT DATA, on how AI is reshaping bank cost structures, operational risk and productivity and why traditional outsourcing models no longer reflect how financial institutions operate today.
For banks facing sustained margin pressure, operational efficiency has become as much a financial concern as a technological one. Processing, servicing and compliance functions account for a significant share of operating costs, while failures in these areas can quickly translate into regulatory sanctions and reputational damage. Rohit Puri maintains that the traditional outsourcing model built around scale and labour arbitrage has reached its limits.
“The biggest shift has been the evolution from labour-arbitrage outsourcing to digitally enabled, AI-powered business transformation,” he says. “Two decades ago, success meant process efficiency and cost control; today, it’s about agility, customer experience, and data-driven insights.”
That shift, he suggests, is changing the underlying economics of bank operations. Rather than optimising individual tasks, banks are increasingly redesigning processes end to end, using AI to improve resilience, control and decision-making — not simply to reduce headcount.
Beyond labour arbitrage
Business process services, long associated with outsourced back-office work, now sit at the centre of how banks manage cost-income ratios and operational risk. Puri is clear that the category itself needs redefinition. “BPS today is about Business Process Orchestration, not just outsourcing,” he says. “It’s about reimagining processes with AI, analytics, and automation to deliver measurable outcomes that matter—customer experience, resilience, and growth.”
For banks, this reframing reflects a broader shift in operating models. Labour arbitrage delivered predictable savings but limited flexibility, particularly as regulatory complexity increased and customer expectations rose. AI-enabled orchestration, by contrast, offers the potential for variable cost structures and greater scalability — provided it is deployed with clear business intent.
“At NTT DATA, we envision ourselves as transformation partners, helping clients build intelligent enterprises that adapt, learn, and deliver impact at scale,” Puri says.
Fixing underperformance: purpose before platforms
Puri has led multiple business turnarounds, and his approach to transformation is notably cautious about technology-led fixes. “I start with purpose and data—understanding why performance is lagging and where value leaks occur,” he says. “From there, it’s about building trust, setting clear metrics, and mobilising teams around a transformation vision.”
Only once that foundation is in place does technology enter the equation. “Technology comes next, layered onto redesigned processes with automation, AI, and analytics,” he says. Just as important, he adds, is culture. “Equally important is culture change—a factor often overlooked but essential for sustainable transformation.”
That sequencing is extremely critical and will transform the way banks approach AI initiatives. “Many organisations start with technology, and not transformation,” Puri says. “Successful AI adoption begins with purpose: define your ‘why,’ rethink processes or data foundations, then align business outcomes, and use AI as the catalyst.”
From efficiency gains to intelligent impact
NTT DATA’s recent work with its BPS Agentic AI Orchestrator and Agentic Factory, known internally as DICE, illustrates how that philosophy is being applied. Puri describes the platform as a way to embed intelligence directly into operational workflows, rather than simply automating existing steps.
“Our BPS Agentic AI Orchestrator & Agentic Factory, also known as DICE, has transformed several complex processes across industries,” he explains. In insurance operations such as policy handling services, “DICE can slash processing times by over 70% and enhance compliance by introducing scalable, autonomous agents to reason, learn, and act autonomously, across workflows.”
Crucially for regulated industries, the system is designed to retain human oversight. “The orchestrator also integrates human expertise through AI Desk – an optional component and brings human in the loop,” Puri notes. The result, he says, is a shift in how work is allocated. “This isn’t just efficiency, it’s intelligent transformation at scale—enabling our talent to focus on judgment, empathy, and innovation, while AI manages precision and speed.”
Why banks are moving first
While adoption of agentic AI is spreading across sectors, Puri sees financial services among the fastest movers. “Banking, insurance, and healthcare are leading the charge in Agentic AI adoption,” he states. The reasons are structural, as Puri points out: “Regulatory complexity, heightened customer expectations, and fierce digital competition are driving these sectors to embrace Agentic AI for compliance, risk mitigation, and hyper-personalised experiences”.
As a result, traditional operating models built around manual, rule-based processes are becoming harder to sustain. “The shift from manual, rule-based operations to adaptive, AI-enabled ecosystems is accelerating,” Puri says. ,.
Preparing people for AI-led operations
The rise of AI is also reshaping skills requirements inside service organisations. Puri says NTT DATA is investing heavily in training to ensure its teams can work alongside intelligent systems rather than be displaced by them.
“We are empowering our talent by investing heavily in upskilling them,” he says. “With nano learnings, instructor led and as well as self-paced learning, that blends domain expertise with digital fluency, design thinking, and ethical use of AI, we are making these trainings mandatory.”
The goal is a shift in mindset. “Teams are trained to view AI as a co-creator, not as a competitor,” Puri says. “This shift in mindset, from execution to orchestration ensures our people remain at the centre of innovation and decision-making.”
The next phase of bank operations
Puri is explicit about what he believes will redefine the sector. “I believe the rise of Agentic AI will completely redefine BPS,” he highlights. He points to NTT DATA’s work on agentic orchestration, digital twins of customers and its IntelACT solution as early indicators of how bank operations could evolve. “These solutions bring self-orchestrating intelligence to processes and systems that can sense, reason, and act autonomously to deliver desired business outcomes,” he says.
Over the next three years, he expects a decisive shift. “BPS will move beyond automation to autonomous transformation where AI continuously optimises operations, ensures compliance, and drives innovation at scale,” Puri notes.
The implication for banks, he concludes, is strategic rather than technical.
“It’s not about doing things faster anymore,” he says. “It’s about doing them smarter, proactively, and with purpose.”
