PhonePe and Razorpay Turn AI Into Tool for Expanding India’s Digital Payments

PhonePe and Razorpay are deploying artificial‑intelligence models to move Indian digital payments beyond simple transactions, aiming to draw an additional 500 million users into the ecosystem and help merchants grow online.

Key takeaways

• AI‑driven insights: Razorpay’s Vulcan 5.2 model analyses roughly 3,000 data points per transaction to predict missing details, flag fraud risk and gauge conversion potential.
• Broader merchant needs: Both firms say accepting payments is only the first step; small and medium businesses also require tools for customer acquisition, retention and conversion.
• Financial‑product expansion: PhonePe’s head of payments, Deep Agrawal, notes the next challenge is offering credit, insurance and other products to existing UPI users.
• Support for SMBs: Razorpay has added payroll, salary disbursement and vendor‑payment services, then expanded to help merchants build and scale online stores.
• Economic impact: Strengthening digital commerce for millions of small firms could boost India’s overall digital‑economy growth.

AI models turn transaction data into actionable intelligence

At TechSparks 2026, Razorpay’s chief product officer Khilan Haria explained that each digital payment carries nearly 3,000 attributes. The company’s proprietary foundation model, Vulcan 5.2, processes these attributes to fill data gaps, assess fraud likelihood and estimate the chance a transaction will lead to a repeat purchase. The output helps merchants tailor offers and reduce checkout failures.

Beyond payment acceptance for merchants

Haria said Razorpay’s early focus on payroll and vendor payments addressed immediate cash‑flow needs of small and medium businesses. Feedback from those merchants revealed a larger problem: attracting demand, improving conversion rates and fostering repeat business. In response, Razorpay is building tools that assist sellers in creating an online presence and optimizing digital commerce pathways.