AWS Powers Razorpay Vulcan: India's First AI Payments Foundation Model 2026

A single AI system trained on nearly 3 trillion data points across 4 billion payments, aiming to make every digital payment in India faster, safer, and smarter.

For millions of shoppers in India, a single failed payment, delayed OTP, or fraudulent card charge can be enough to send them back to cash. As India's e-commerce market heads toward a projected $350 billion by 2030, the reliability of digital payments becomes a national economic priority. Razorpay, in collaboration with NVIDIA and AWS, has responded by building a unified AI foundation model that scores every payment route in real time and picks the healthiest path before the transaction is even attempted.

What was announced

Razorpay announced the launch of Razorpay Vulcan, India's first Transformer-based AI foundation model built specifically for payments. The model combines Razorpay's payments data with NVIDIA's accelerated computing and AWS cloud infrastructure, including Amazon SageMaker, to deliver a single continuously learning intelligence layer for the payments ecosystem.

Ahead of the full launch, early components of the model have been running across 3 trillion data points on Razorpay's network, testing routing, fraud, and risk decisions on live transactions. Customers including Blinkit, Bachatt, and redBus have already reported measurable results:

• Payment success rate: 8-10% improvement
• International card fraud: 8x more fraud detected and stopped
• Fraudulent or disputed transactions: 5x more identified without increasing alerts
• Checkout personalisation: 40% more shoppers see their preferred UPI app, helping complete 1-2 lakh additional purchases every month

Why this matters

For CIOs, CISOs, and IT leaders in fintech, retail, and financial services, Vulcan signals a shift from fragmented specialised ML models toward unified foundation models that share signals across routing, fraud detection, risk scoring, and checkout. Instead of maintaining and retraining separate systems, enterprises can rely on one model that learns continuously from every transaction.

Razorpay is reimagining payments intelligence at India scale with an AI Foundation Model built on Amazon SageMaker that consolidates billions of transaction insights into a single, continuously learning intelligence layer, replacing fragmented ML models with unified AI that delivers higher payment success rates, rapid iteration, and enterprise-grade security for mission-critical payment flows.

Technical details

• Architecture: Transformer-based foundation model adapted for payments data patterns
• Training scale: Approximately 3 trillion data points across 4 billion payments
• Signals per transaction: Roughly 3,000
• Compute: NVIDIA GPUs for training and large-scale inference
• Cloud platform: AWS infrastructure with Amazon SageMaker for development, training, and deployment
• Capabilities: Hyper-precision routing, network-level fraud detection, RTO risk intelligence, predictive checkout personalisation
• Ownership: Proprietary architecture and training data belong to Razorpay

Softprom and Amazon Web Services

Softprom is the official partner of Amazon Web Services. Our team helps enterprises design, migrate, and scale AI and data workloads on AWS, including services such as Amazon SageMaker for machine learning and generative AI initiatives.

This content was prepared as part of the Softprom DistriFlow project — an automated system for monitoring and adapting vendor news. Original source: original article.