Vijay Vijayasankar is a Global Agent AI Officer at Genpact. He is responsible for driving Genpact’s agent AI vision and accelerating innovation through strategic partnerships across data, AI, go-to-market and delivery capabilities. Prior to joining Genpact, Vijay held senior management positions at IBM, MongoDB, and SAP. He holds a degree in Mechanical Engineering and an MBA from the University of Kerala. During my interaction with TOI, vijay vijayasankar We shared Genpact’s AI-driven efforts and more. Genpact says it has moved beyond traditional BPO to “AI-driven autonomous operations,” but what does this change mean in terms of revenue mix, and how much of today’s business is already AI-driven?At Genpact, the transition to AI-driven autonomous operations is focused on embedding AI into the fabric of how work is done, rather than isolating AI as a revenue stream. Our advanced technology solutions business, which spans data and AI, digital technology, advisory and agent solutions, continues to grow and now represents a significant proportion of our overall business. It’s not a trend. This is a clear signal of where a company’s demand is heading. More broadly, almost half of our portfolio has AI in place, highlighting the central role it plays across our clients. What is changing in the revenue mix is the nature of the engagement itself. Rather than traditional BPO with people-based contracts, business customers are increasingly demanding outcomes such as touchless invoice processing with our agent accounts payable suite and faster close processing with our agent record-to-report suite. You are positioning yourself to build “agent AI systems” rather than providing a service. How has the way you interact with clients changed compared to traditional outsourcing arrangements?With 30 years of extensive experience running G&A and COGS functions for the world’s leading companies, Genpact understands the nuances of the last mile that make or break each process. This allows us to offer our clients productized agent AI solutions that are different from traditional AI and software products and traditional service contracts. We deliver data, process, and operator experience knowledge codified into agent solutions at scale to support business outcomes. The conversation is no longer about worker arbitration, but about how humans can intervene in judgment and exception handling to intelligently coordinate work and execute it at scale. And the market is responding. 2025 saw significant adoption of Genpact AP Suite, our first suite of agent solutions for accounts payable. Of particular importance is that a significant portion of that demand comes from net new customers. It is not a rotation of existing business books. This is evidence that the total addressable market is expanding significantly. For existing clients who have switched from an FTE-led model to an agent model, we are seeing revenue growth and gross margin expansion that exceeds our own projections starting in mid-2025. As AI takes over repetitive processes, what roles are actually increasing within Genpact and are we seeing a shift towards data, AI training, and domain-driven roles? The easiest way to explain what’s happening is that everyone at Genpact is now an AI practitioner, and some of our colleagues are AI builders. Systems are designed by builders such as data engineers, AI architects, and agent solution developers. Practitioners are experts in their field. Finance executives, supply chain specialists, and risk analysts now work with agents rather than spreadsheets. Now I would like to return to the broader story. The hopeless frame that AI is destroying jobs misunderstands that AI is actually good. Let’s take programming as an example. This is perhaps the single most powerful use case for AI today. The number of software engineering jobs is not decreasing. If anything, they’re getting better. why? Because actually typing code is probably 10-20% of a software engineer’s job. Mindset, architecture, problem definition, that’s the other 80%. AI takes over the repetitive parts and allows humans to do the more valuable parts. The same logic applies to all areas we work on. The people who are at risk of losing their jobs are those who resist using AI to redefine their roles.As work is redefined, people are being trained and redeployed to work alongside AI agents. Supply chain managers who six months ago were writing reports are now managing agent workflows and interpreting exception flags generated by AI. Will this transition mean less growth in employee numbers compared to traditional BPO models, or will it create new categories of jobs within the organization?The linear relationship between revenue growth and employee growth is breaking down, but this is by design rather than chance. In traditional BPO models, growing your business by 10% requires adding 10% of your headcount. Agentic Operations completely changes that equation. At Genpact, the composition of our workforce is changing more than our overall size. We continue to aggressively invest in our AI talent, both through hiring technology experts and intentionally training and upskilling our teams. Clients are increasingly looking for results, not people. Are contracts now moving to an AI-driven, outcome-based pricing model? And how does that impact margins?Yes, the outcome-based model for agent operations has increased margins and clients get better results. In this way, the model is designed from scratch. Share an example that resonates with your clients. Think about the civil society in which we live. We need laws and law enforcement to protect it. Both local and state police officers operate within defined jurisdictions. If a suspect crosses state lines, police cannot track him. However, federal agents with broader jurisdiction (such as the FBI in the United States or the CBI in India) can do so. This is how agent operations work. AI agents handle what I call “happy flows” – well-defined and repeatable tasks within a jurisdiction, such as standard invoice matching. If something is outside of its jurisdiction, highly trained humans, or “special agents,” intervene. The key insight is that you can’t just send in federal investigators for a simple noise complaint. Because rank-and-file police officers are required to protect 80% of their jobs, the most skilled officers are reserved for truly necessary exceptions. That’s the model. As AI becomes an execution layer, client expectations will naturally become more about results than effort. This coincides with a broader shift to AI-first operations, where value is delivered through precision, speed, and scale powered by AI rather than human resources.From a distribution perspective, this allows for more predictable and efficient operations with straight-through processing and reduced volatility. Over time, this will support the transition to a value-linked engagement model where price becomes increasingly tied to business outcomes. From a margin perspective, the benefits are clear as the model becomes more efficient and scalable, even as we continue to invest in our strategic priorities. The shift to non-FTE, outcomes-based delivery is a key driver. In our advanced technology solutions business, the majority of our work is repetitive and delivered through non-FTE models, combining durability with the economics of high-quality at scale.As global companies combine IT and BPM capabilities, do you think the industry is moving towards a fully integrated model of AI, consulting, and operations?The industry is clearly moving toward a model where AI, consulting, and operations converge into one integrated function. The definition of success is the guaranteed result, not the value of the time spent building it.What distinguishes leaders is not just their ability to deploy AI, but how they intentionally engineer autonomy at scale across data, architecture, orchestration, and governance. I think one of my roles is really to drive this integration within the company. The Global Agent AI Officer role exists because organizations are in transition from a primarily human-centric problem-solving approach to a more autonomous enterprise model. We bring technology and processes into the same room, connecting deep domain and industry knowledge, proprietary data, and agent systems to truly integrate AI and transform your business.
