Inside Genpact's Engine Room: The Future of Agentic AI

Understanding the Engine Room

Balkrishan “BK” Kalra, the CEO of Genpact, emphasizes that the company's key innovations are not in the showroom but right in the engine room—where actual operations such as finance, procurement, and supply chain are managed. For nearly thirty years, Genpact has been a hidden force behind the operations of major corporations, initially under General Electric, later emerging as a leading business-process services company.

Now, Kalra is poised to convert this operational expertise into a new paradigm for AI services, focusing on what he calls agentic operations.

The Agentic Operations Model

The crux of Genpact's strategy lies in the integration of AI with deep operational insights. Kalra articulates that enterprise AI isn't merely about enhancing models; it's about embedding process intelligence—understanding the nuances of decisions and exceptions that occur within business workflows.

“That distinctive domain is shining more,” said Kalra, reflecting on the moment Genpact is capitalizing on.

Under Kalra's leadership, Genpact's model allows AI agents to execute tasks while human experts oversee exceptions and ensure adherence to ethical AI practices. The company is applying this model across various operational domains, including accounts payable and insurance.

Standing Out in a Crowded Market

In a saturated market, Genpact differentiates itself by focusing not just on superior AI access but on applying it effectively within actual workflows. Kalra stresses the importance of contextual knowledge—understanding how enterprise work operates, where it falters, and how those hiccups can be rectified.

The Reality of AI Deployment

However, deploying AI in enterprise environments is not a straightforward technology rollout. As AI systems grow increasingly complex, the challenge lies in managing the delicate balance between machine processing and human oversight. The goal is to enhance human roles rather than replace them.

Vijay Vijayasankar, Genpact's global agentic AI officer, underscores that for AI to be valuable, the surrounding architecture needs to be robust. Raw AI models often fall short of expectations in enterprise contexts, and many business cases don't necessitate the use of complex models.

Facing Enterprise Debt

A report from HFS Research, in conjunction with Genpact, reveals that companies encounter various forms of “enterprise debt,” including data and talent deficits which inhibit the realization of AI's full potential. The study indicates that a staggering 86 percent of business leaders acknowledge this impediment, and only a small fraction have taken effective steps towards remediation.

Data shows these debts are not fleeting issues; they could unlock nearly $18 trillion in enterprise value, suggesting that the need for innovation is now more pressing than ever.

The Dual Velocity Approach

This dual-velocity approach enables companies to work on immediate AI applications while concurrently addressing the root issues in technology, data management, and organizational processes. Companies that thrive will not treat AI as mere projects but will integrate it into core business functions.

“AI is just sort of there. It's the underpinning of everything,” notes Saurabh Gupta, president of Research and Advisory Services at HFS Research.

Preparing for the Future Workforce

Kalra firmly believes that employees need to evolve into either AI builders or practitioners. This emerging workforce will have to bridge the gap between cutting-edge technology and traditional business domains like finance and supply chain.

Conclusion: The Future of Enterprise AI

The ongoing evolution of agentic AI exemplifies a test for large enterprises: Are they capable of redesigning their workflows quickly enough to harness the technology they are investing in?

As Kalra puts it, while Genpact may not present as the most glamorous player in the tech landscape, it recognizes its role as a pivotal engine-room entity, focusing on enhancing core business functions—from invoices and claims to procurement approvals. “We enable the showroom,” he asserts, encapsulating Genpact's mission to transform enterprise AI for the better.