IPO-bound digital payments company Razorpay is betting on artificial intelligence (AI) to fuel its next phase of growth.
The Bengaluru-based fintech startup plans to use agentic AI in commerce, which allows systems to act autonomously for businesses or customers, to help small and medium enterprises (SMEs) process transactions faster and more smoothly.
“We’re leaning hard into AI. It’s creating the largest change at the moment,” Harshil Mathur, chief executive and co-founder at Razorpay, told Mint on the sidelines of the fintech firm’s annual FTX summit. “What we’re trying to figure out is to how to bring the power of AI to the small and medium enterprises that we power in the country.”
Agentic commerce is becoming increasingly popular as brands and platforms try to figure out how to make the shopping experience for customers more seamless and intuitive. At the same time, firms like Razorpay are attempting to build the backend payment and orchestration systems that these companies will use to power their businesses.
Agentic AI systems have been the darling of investors through 2025 for multiple reasons. In fact, the promise behind the technology is that it can adapt and learn on its own to the point that businesses can run several agents that talk to each other for processes that are largely routine and currently use manual labour.
For Razorpay, its agentic AI suite, announced on Thursday, marks a shift for the company. “It transitions us from being just a payments provider for SMEs to where we become the financial operating system for them,” Mathur said referring to SMEs.
Market opportunity
Razorpay is increasingly focused on making it easier for SMEs to do business and scale faster. Many of these enterprises face hurdles such as fraud, failed payments, and difficulty tracking vendors, which can slow their growth.
Razorpay says that it wants to solve all these problems by letting its users create agents that can field these problems independently, instead of having to hire someone for these tasks. “Businesses don’t want to manage the complexity of dealing with multiple platforms,” said Mathur. “It’s why it is an opportunity for us because they’re already doing a lot of their payments systems on our platform.”
There are currently 79 million registered MSMEs in India, according to the Udyam portal run by the Union ministry of micro, small & medium enterprises. The sheer size of the market makes it an opportunity for Razorpay to disrupt and add another stream of revenue to its current business.
The fintech believes that eventually the AI suite will account for a large chunk of its business, but Mathur said that it was too early to tell how much. “Adoption is still adoption. Maybe customers figure out their own way of solving problems that don’t need us. We’ll have to see how it plays out in reality.”
While the company may have launched an AI suite meant to ease business, investments like these tend to be expensive, both from a tech and monetary standpoint. But, for the time being, Razorpay is not charging anything for the use of the suite. “When adoption builds, we’ll think of a way to charge. There’s enough global examples of how you typically charge for agents, where they’re free up to a limit and then you charge for tokens. We’ll adopt something similar, but it’s too early to say,” Mathur added.
