Summary

Razorpay has built an AI payments foundation model — dubbed "Vulcan" — trained on four billion transactions to improve payment success rates, fraud detection, and checkout personalisation as India's digital payments ecosystem scales. Built with NVIDIA and AWS, the model uses around 3,000 signals per transaction to analyse payment flows across merchants, banks, instruments, and gateways. Early tests across 1.5 million transactions and more than 51,000 businesses showed an 8-10% improvement in payment success rates, an eight-fold increase in international card fraud detection, and a five-fold increase in identification of fraudulent or disputed transactions.

Key Facts

Why It Matters

Razorpay's "Vulcan" is a landmark for AI-native payments — a single multidimensional foundation model that can handle success-rate optimization, fraud detection, and personalisation, replacing siloed ML models. It shows India building its own payments AI rather than importing models designed for other markets, reflecting the country's distinct UPI/card/net-banking architecture. As India's digital payments scale, foundation models that understand "the language of money" could become core infrastructure — and a monetisation lever for Razorpay via volume growth and lending, even as it pursues an IPO.

Sources

Powered by Forestry.md