Thursday, March 12, 2026
HomeStart UpRazorpay builds AI payment agents with Claude to automate payments

Razorpay builds AI payment agents with Claude to automate payments

Razorpay has announced the development of AI-powered payment agents built using Claude, marking a significant step toward integrating artificial intelligence into digital payment operations. The company revealed that these intelligent agents can automate routine financial tasks such as recovering abandoned purchases, retrying failed subscription payments, resolving disputes, and forecasting cash flows.

Furthermore, the initiative reflects a broader transformation in the payments industry toward agentic commerce, where AI systems actively perform financial and operational tasks on behalf of businesses. As a result, payment companies are increasingly exploring ways to allow customers to complete transactions directly through AI assistants and conversational interfaces.

Several players across the payments ecosystem are also moving in this direction. Payment aggregators such as Cashfree and global card networks Visa and Mastercard, as well as merchant processors including PayU and Pine Labs, are working with advanced AI platforms such as ChatGPT and Claude to enable similar capabilities.

Razorpay integrated these AI agents through Anthropic’s Claude agent software development kit, enabling businesses to deploy intelligent systems that interact directly with customers. For example, the AI agents help merchants recover lost sales by contacting shoppers who abandon their online carts. The agents can send a message or voice notification to ask why the purchase remained incomplete and then offer reminders or small incentives to encourage customers to finish the transaction.

In addition, the fintech company has launched an agentic experience platform, which introduces an AI-native layer designed to simplify the way online businesses onboard to Razorpay, integrate payment infrastructure into their applications, and manage payment workflows more efficiently.

“Businesses don’t just need more software anymore, they need intelligence that can act,” Harshil Mathur, chief executive of Razorpay, said in the statement.

Moreover, the company stated that businesses can use these tools to create their own customized AI agents through simple language-based commands. These agents can integrate seamlessly with e-commerce platforms such as Shopify and logistics platforms like Shiprocket, while also connecting with messaging services such as WhatsApp.

Irina Ghose, managing director, India at Anthropic, said, “Razorpay’s work with Claude shows how AI agents can recover revenue, resolve disputes, and predict cash flow. It’s a great example of what AI can do when it is built into business operations.”

At the same time, Razorpay is experimenting with AI-led in-app commerce experiences across several consumer platforms. The company is currently testing these capabilities with partners including Zomato, Swiggy, PVR INOX, and Vodafone Idea.

Through these experiments, customers can discover products, evaluate options, and complete payments within the same AI-powered conversation. Consequently, the approach could significantly streamline digital commerce by reducing the steps required to finalize transactions.

Razorpay’s move to build AI-powered payment agents highlights the rapid evolution of fintech toward intelligent, automated commerce systems. By integrating conversational AI directly into payment operations, the company aims to help businesses recover lost revenue, simplify financial management, and enhance customer engagement. As AI-driven commerce gains traction across the payments ecosystem, innovations like these could reshape how businesses and consumers interact with digital transactions.

Subscribe To Newsletter

ICYMI

BRL Editor
BRL Editorhttps://businessreviewlive.com
Business Review Live covers finance, technology, travel, lifestyle, and everything in between through exclusive interviews and analysis, market statistics, digital video, and an expanded array of content formats.