Razorpay has introduced Vulcan, claimed as India’s first AI foundation model focused on payments, designed to enhance transaction reliability and reduce fraud by learning from trillions of data points across billions of transactions.

  • Vulcan is trained on 4 billion payments and 3 trillion data points.
  • Developed with Nvidia and AWS, used by Blinkit, Bachatt, and redBus.
  • Legal and data consent questions remain unresolved.

What happened

Razorpay launched Vulcan, described as India’s first transformer-based AI foundation model specializing in payments. The model handles critical aspects such as payment routing, fraud detection, risk evaluation, and personalization. Razorpay collaborated with Nvidia and Amazon Web Services to develop and host this model natively within India.

Early testing is underway with companies like Blinkit, Bachatt, and redBus integrating Vulcan’s capabilities into their live payment flows. The model is designed to leverage a vast dataset of approximately 4 billion payments representing 3 trillion data points, offering a unified approach compared to multiple task-specific AI models.

Why it matters

By training a single AI foundation model on a comprehensive dataset of payment transactions, Vulcan aims to significantly improve the consumer payment experience by reducing failed transactions, lowering fraud-related losses, decreasing undelivered orders, and boosting checkout success rates. Razorpay positions Vulcan as a key technology to foster greater trust in digital transactions among Indian users still weighing cash against digital payments.

However, several important considerations remain. Razorpay has not disclosed detailed methodology or training data specifics, including whether consent from individual consumers or merchants covers the use of transaction data for AI model training. Given its role as a data processor under India’s Digital Personal Data Protection Act, the legal basis for this training is unclear. Moreover, plans to extend Vulcan into authentication and lending introduce potential fairness and transparency challenges, especially in credit decision-making.

What to watch next

Observers should monitor how Razorpay addresses outstanding questions about data usage consent and compliance with data protection regulations, particularly as it expands Vulcan's role from payments to lending and authentication. The ethical implications of AI-driven credit assessments, including explainability and non-discrimination, will be key areas of scrutiny.

Additionally, with Razorpay preparing for a public listing supported by this technology, stakeholders will be interested to see how Vulcan evolves as a commercial asset powered by continuous real-world transaction data. The success of this model could influence how other Indian fintech companies adopt AI in their payment and financial services infrastructure.

Source assisted: This briefing began from a discovered source item from MediaNama. Open the original source.
How SignalDesk reports: feeds and outside sources are used for discovery. Public briefings are edited to add context, buyer relevance and attribution before they are published. Read the standards

Related briefings