Why Indian enterprise could lead adoption of agentic AI

India stands at a unique inflection point in the global technology landscape. In 2026, the conversation has shifted from the mere potential of Artificial Intelligence (AI) to its practical and autonomous execution. The country has one of the world’s largest developer workforces and a proven history of digital leapfrogging, which has been seen previously in the UPI and digital payments revolution. India is now primed to lead the shift toward the Agentic Enterprise.

In the latest episode of Mint's series All About AI, global and regional leaders from Salesforce came together to discuss what this evolution looks like. The consensus is that the era of “pilot purgatory”, where AI projects remain stuck in experimental stages, seems to be coming to an end.

Moving beyond the last mile

For many CEOs, the initial excitement surrounding AI models has been met with a value gap. While models are performing exceptionally well in evaluations, bringing that value to the corporate bottom line has remained a challenge. This is often described as the “diffusion gap”.

Srini Tallapragada, President and Chief Engineering and Customer Success Officer at Salesforce, said that an agentic enterprise requires more than just a smart model. It needs a foundation of trust and context.

“For an agentic enterprise to really deliver, you need a trusted context layer at the bottom. Agents need context, a way to understand system workflows, and a system of agency that brings it all together with governance and security,” he said.

Tallapragada added that 2026 is the year where AI must show tangible business outcomes. Salesforce itself has acted as “Customer Zero”, reportedly saving over $100 million on the bottom line through customer success use cases powered by these autonomous agents.

Solving scale with intelligence in India

India’s position in this shift is not just as a consumer, but as a primary orchestrator. During the digital transformation era, the country used technology to fix legacy challenges. Today, AI offers a way to solve problems at a scale that was previously unimaginable.

Arun Kumar Parameswaran, EVP and MD of Salesforce South Asia, felt that while digital transformation focused on digitising existing processes, the agentic era is forcing a total reimagination of business models.

“Our ability to solve problems at scale is exponentially increased with AI. We are not just doing a lift and shift. We are redefining how customers do business, changing their access to market and their entire experience standpoint,” said Parameswaran.

He further highlighted that India’s demographic and existing digital infrastructure including the India Stack, catapulted the nation to the top ranks of the digital economy. AI is expected to follow a similar trajectory, not just as a tool for productivity but as a teammate that can handle millions of citizen or customer interactions simultaneously.

Bridging the value gap with data

The transition to an agentic model isn’t that simple. Tallapragada said that no AI transformation can begin without a thorough data transformation. Many enterprises still struggle with federated data with information scattered across different silos.

The last mile of AI adoption often requires fixing these “messy” brownfield environments. This is where the Indian System Integrator (SI) ecosystem, including giants like TCS and Infosys, plays a critical role. They provide the human expertise needed to navigate complex legacy systems and integrate autonomous agents that can “think” and “act” within the business context.

Defining the three pillars of success

For organisations looking to make this transition, the leaders suggest a focused three-prong strategy:

• Pick the right partner: “Do-it-yourself” is viable for initial coding, but managing the operation from day two onwards requires a robust platform. As Tallapragada put it, “Everybody loves vibe coding. Nobody loves the vibe operating.”
• Focus on the business case: Instead of starting with the technology, start with the hardest business problem. Whether it is top-line growth or bottom-line efficiency, the outcome must be clear.
• Prioritise change management: The shift to agentic AI is as much about people as it is about software. Organisations must provide the psychological safety for teams to learn, iterate, and adapt to new ways of working.

The “Build and tear down” mindset

Perhaps the most radical shift for Indian enterprises in 2026 is the pace of obsolescence. In the old world, systems were built to last a decade. In the agentic world, the baseline assumption is that whatever is built today may need to be torn down and improved within 12 months due to the exponential pace of innovation.

The cost economics in India present a unique challenge that human agents are often cheaper than AI tokens. However, forward-thinking companies are investing ahead of the curve, recognising that the competitive advantage of being AI-first outweighs the immediate cost-benefit analysis.

The roadblocks, whether they be power constraints for data centres or the need for reskilling, are viewed by leadership as enablers rather than blockers. The final advice for Indian leaders is simple – stop the endless introspection and start doing.

“A lot of anxiety goes away when you start using it,” Tallapragada said. By picking small projects and implementing them, enterprises can create the feedback loops necessary to win in the agentic era.

Note to Readers: This is a Mint editorial initiative, sponsored by Salesforce.