India-based AI company Gnani.ai has partnered with fintech firm Razorpay to launch an agentic AI platform that enables payment collections to be completed during live customer calls.
The announcement was made on 26 February 2026 and is aimed at addressing limitations in existing automated debt recovery systems.
Traditional voice AI systems can prompt customers to make payments, but typically require them to complete transactions through a separate channel, a process the companies say results in lower conversion rates. The new platform allows AI agents to initiate and settle payments within the same conversation, without human intervention.
How the integration works
The platform connects Gnani.ai's voice automation technology with Razorpay's Model Context Protocol (MCP) Server, described as an AI-native payments layer that enables software agents to initiate, authorise, and monitor transactions in real time. By linking directly to Razorpay's infrastructure, the AI can generate payment links or UPI requests, track completion status, and confirm settlement within the call.
The platform currently supports one-time UPI payments and mandate-based recurring payments. Support for cards and digital wallets is planned for future integration.
Ganesh Gopalan, Founder and CEO of Gnani.ai, said the integration is set to deliver what voice AI has long promised but not achieved, namely the ability to complete transactions rather than simply initiate conversations, allowing enterprises to automate collections at scale without sacrificing conversion.
Khilan Haria, Chief Product Officer at Razorpay, noted that agentic AI only becomes valuable when it can act rather than just converse, and that the partnership gives AI systems a secure, real-time pathway to complete payments inside live customer interactions.
Scale and language support
Gnani.ai states its system processes more than ten million calls daily and supports over 38 languages, including Hindi, Tamil, and Bengali, enabling deployment across varied customer bases in India and beyond. The breadth of language support is a relevant factor in the Indian market, where linguistic diversity has historically been a barrier to automated customer engagement at scale.
