Razorpay Launches AI Agent Studio to Automate Business Operations

Razorpay has launched Razorpay Agent Studio, a pioneering platform that integrates autonomous AI agents directly into payment infrastructure to automate operational tasks. The platform allows businesses, especially small ones, to deploy agents for functions like dispute management, customer recovery, and payment reconciliations. It also introduces "agentic commerce," enabling users to make purchases through chat interfaces within apps like Zomato. CEO Harshil Mathur emphasized the platform's ease of integration and its built-in AI security agents that monitor every transaction for fraudulent patterns.

Key Points: Razorpay AI Agent Studio Automates Payments & Business Tasks

  • Automates dispute management & reconciliations
  • Enables "agentic commerce" via chat
  • Integrates in under two minutes
  • AI monitors every transaction for fraud
3 min read

"World's first platform built on top of payments": Razorpay CEO Harshil Mathur on AI Agent Studio for autonomous operational tasks

Razorpay launches AI Agent Studio, a platform built on payments to deploy autonomous AI for dispute management, fraud prevention, and agentic commerce.

"The business of running a business can completely go away because a single person can operate like a team of 100 agents. - Harshil Mathur"

New Delhi, March 12

Razorpay, on Thursday, announced the launch of Razorpay Agent Studio, a first-of-its-kind platform built on top of payments that allows businesses to deploy autonomous AI agents for operational tasks. The platform aims to automate various business processes, including dispute management, customer recovery, and payment reconciliations, by integrating AI directly into the payment infrastructure.

Harshil Mathur, CEO & Co-Founder, Razorpay, speaking to ANI at the sidelines of the FTX 2026 event, said, "This is the world's first platform. We have an agent platform that is built on top of payments."

He noted that while large companies have teams to manage such tasks, small businesses often do not.

"So they can spin up agents on the fly and let them do all of this stuff for them. It can be something simple like card abandonment recovery. It can be dispute management. It can be any number of things that a customer wants to do," he said.

According to Mathur, the shift toward AI agents transforms the efficiency of small enterprises. "The business of running a business can completely go away because a single person can operate like a team of 100 agents," he said.

He explained that these agents can autonomously handle disputes by checking payment status, order platforms like Shopify, and third-party logistics providers for proof of delivery. "The entire dispute management process can become automated," Mathur noted.

The company has also introduced "agentic commerce," allowing users to make purchases through chat interfaces within existing apps.

Mathur explained how a user could interact with the Zomato app to place an order autonomously. "You can chat with ChatGPT, you can chat with Claude and you can buy things. Now we're extending that to inside apps," he said.

"You can open the Zomato app, chat with it and say, 'Hey, I want samosa from this, I want chai from this,' and the Zomato app can buy it for you. Completely autonomous just by chatting on it," Mathur noted.

On the technical side, Razorpay has simplified the integration process for developers. "The human part of writing code, reading documentation, completely goes away. We demonstrated live today that you can complete integration in less than two minutes," Mathur said. He added that the system is self-evolving, noting, "If a new instrument gets added, you just have to give one more prompt, and the system will take care of it."

Regarding security, Mathur emphasised that AI agents now monitor every transaction to prevent fraud. "Every single transition on Razorpay is now monitored by an AI agent, which is able to figure out what kind of transition is this, does this pattern match with other kind of fraud transitions we've seen, and highlights it and blocks it before it even happens," he stated.

He clarified that financial data remains secure as "the agent never sees any of that information" and transactions only occur within "the guardrails of your consent."

- ANI

Share this article:

Reader Comments

A
Arjun K
The "agentic commerce" part is fascinating. Chatting with Zomato to order samosa and chai? That's the future. But the security part is crucial. Glad they addressed it, but I'll wait for real-user reviews before trusting an AI agent with my transactions.
R
Rohit P
"A single person can operate like a team of 100 agents" – this is the promise of AI that excites me. For a country with so many young entrepreneurs, tools like this can level the playing field against bigger companies. Razorpay is doing some innovative work.
S
Sarah B
The two-minute integration claim is impressive if true. Developer time is expensive. Automating the coding part for payment setups could save startups a lot of money and speed up their launch. Cautiously optimistic.
V
Vikram M
While the tech sounds cool, I have a respectful criticism. We must be careful not to automate away the human touch entirely, especially in customer recovery. An AI message might not handle an upset customer from a Tier-2 city the same way a thoughtful person could. Efficiency shouldn't come at the cost of empathy.
K
Karthik V
Fraud detection by AI on every transaction is a massive step forward. UPI fraud is a real concern. If this can reduce it significantly, it will build more trust in digital payments. Bharat needs this kind of security innovation. 👍

We welcome thoughtful discussions from our readers. Please keep comments respectful and on-topic.

Leave a Comment

Minimum 50 characters 0/50