Latest news with #Pulyani

Business Standard

Cashfree Payments rolls out AI-first module across core products

Fintech firm Cashfree Payments has rolled out a beta version of a software module that integrates large language models (LLMs) and the company's own application programming interfaces (APIs) to improve efficiency for merchants. The latest piece of technology, also known as a model context protocol (MCP), automates several processes across use cases such as payments, payouts, and verification. At present, Cashfree Payments and Razorpay have launched the offering for their merchant base. The MCP server is capable of processing natural language queries while utilising the Bengaluru-based company's proprietary APIs to trigger actions such as generating payment links or verifying Know-Your-Customer (KYC) credentials. It is targeted at merchants of all sizes. For instance, integrating an MCP server into a merchant's system can automate the generation of payment links—a process that typically takes around 15 minutes per order—reducing it to just a few seconds. 'From a merchant's point of view, you just have to install a single MCP server and your entire job is done. Merchants can go live within 10 minutes with the offering. We have not commercialised the MCP server yet,' said Nitin Pulyani, SVP, Product, Cashfree Payments, in an interaction with Business Standard. Pulyani explained that the company has adopted the service across its core products such as payment collections and KYC, and has a proof of concept in place for the payouts feature. To make this server possible, Pulyani noted that the company had tested multiple models at a time when they were evolving rapidly. 'There's evolution happening across the ecosystem. Based on that, our usage, trial, and experience with each of these models is also evolving. We worked with Claude, Anthropic, Gemini, and to some extent with Grok too to write the code,' he said. He added that most of the models currently available are mature enough to operate at the scale required by a company like Cashfree. 'Capacity will not be a constraint. With the pace of evolution and faster adoption, usage will grow significantly. The efficiency of businesses will go up eventually,' he said. The integration of an MCP with Cashfree's APIs and a merchant's dashboard enables the module to identify the context of an operation and perform an action based on this contextual understanding.