Razorpay has announced a partnership with the Bengaluru-based startup Sarvam AI to create voice-first, conversational commerce experiences in India. The collaboration leverages Sarvam’s AI models and agentic stack alongside Razorpay’s payment infrastructure, allowing users to browse products, place orders, and complete payments through natural voice commands in multiple Indian languages.

The system is designed to understand user intent and manage the entire transaction flow without relying on traditional app navigation. The initial rollout will feature Swiggy on the Indus App, enabling users to order food simply by speaking to an AI assistant. According to a CNBC report, businesses can also integrate these voice-based commerce features into their own platforms. As part of an early deployment, a conversational assistant has already been implemented on The Derma Co website, allowing customers to explore and purchase products using voice commands.

Sarvam’s technology will also be integrated into Razorpay’s Agent Studio, giving developers the ability to build multilingual AI agents capable of interacting in languages such as Hindi and Hinglish. The goal of this collaboration is to make digital commerce more accessible to India’s multilingual population, with AI agents handling everything from product discovery to checkout in a seamless conversational flow.

Last month, Razorpay also partnered with Indian AI startup Gnani.ai, which focuses on a more specific application compared to the broader conversational commerce initiative with Sarvam AI. Together, they launched an agentic AI collections platform that helps businesses complete payment transactions during live customer calls. The AI agent can assess intent, generate payment requests (like UPI links), and confirm payments within the same interaction.

Unlike the Sarvam AI partnership, which covers end-to-end conversational commerce from discovery to checkout, the Gnani.ai platform is targeted at automating payment collections. It integrates Gnani.ai’s voice AI with Razorpay’s payments infrastructure to manage the full payment workflow in real time, including verification, link generation, tracking, and confirmation, focusing on financial operations rather than general consumer transactions.

Sarvam AI is a Bengaluru-based startup developing speech, language, and multimodal AI systems specifically for Indian use cases. Rather than a single general-purpose chatbot, the company builds specialized models for tasks like speech recognition, text-to-speech, translation, and document understanding, with a strong emphasis on Indian languages. Its portfolio includes Saaras (speech recognition), Bulbul (text-to-speech), Saarika (transcription), Mayura (translation), and Sarvam-M (multilingual reasoning model). On the vision front, Sarvam Vision handles OCR and document analysis, while applications like Samvaad enable voice-based interactions.