The fintech firm is shifting toward agentic service models to automate payment operations and recover significant gross merchandise value for merchants.
Cashfree Payments is integrating autonomous artificial intelligence agents into its merchant infrastructure to automate critical payment operations and recover lost transaction revenue. This strategic move marks a shift for the Bengaluru-based firm as it seeks to move beyond traditional payment processing toward high-margin, agentic service models.
The platform, branded as Relay, functions by executing specific operational tasks that previously required human intervention. These capabilities include retrying failed payment attempts, managing subscription failures, and resolving transaction disputes. Merchants can also utilize the system to verify cash-on-delivery orders before dispatch, thereby reducing logistics overhead.
A primary objective of the deployment involves reclaiming lost gross merchandise value, which the company estimates at ₹20,000 crore. The AI agents interact with customers via voice or text channels to address friction points during the checkout process. By offering real-time incentives or clarifying payment errors, the system aims to convert abandoned carts into completed transactions.
Mayank Juneja, director of engineering at Cashfree Payments, noted that these functions were historically managed by dedicated growth teams. These teams would manually contact customers to identify reasons for abandonment or payment failures. The transition to automated agents allows these businesses to reallocate human capital toward more complex strategic initiatives.
Operational efficiency gains represent a significant value proposition for small and medium businesses. Cashfree reports that a typical merchant spends approximately 60 hours per week on payment-related operations. The company claims its automated agent can compress this workload to roughly 45 minutes, representing a substantial reduction in administrative burden.
Data privacy remains a central component of the firm’s architecture. Cashfree stated that merchant transaction data is processed entirely within its proprietary infrastructure. This design choice avoids the common practice of transmitting sensitive financial information to external third-party artificial intelligence providers.
Beyond the retail sector, the utility of these agents extends into financial services and lending. Juneja indicated that the technology is being adapted for loan recovery workflows. This involves analyzing e-payment mandates and determining optimal contact windows to engage customers regarding outstanding balances.
The introduction of this service creates a new revenue line for Cashfree in a market characterized by thin margins. While the initial rollout is free, the firm intends to transition to an outcome-based pricing model. Co-founder Reeju Datta explained that fees will eventually be tied to measurable results, such as the volume of recovered carts and the number of successfully resolved disputes.
This development arrives as the broader Indian fintech sector accelerates its adoption of specialized artificial intelligence models. Competitors are also investing heavily in infrastructure to enhance transaction success rates and fraud detection. For instance, Razorpay recently launched an integrated model capable of analyzing thousands of signals across billions of transactions to optimize payment routing.
The shift toward agentic workflows suggests that payment providers are increasingly positioning themselves as operational partners rather than mere utility conduits. By embedding intelligence directly into the checkout flow, these firms are capturing more value from the transaction lifecycle. This evolution reflects a broader trend where automation is used to solve specific, high-friction problems in the digital economy.
The reliance on these agents highlights a growing necessity for firms to differentiate their offerings through proprietary data processing. By training models on internal transaction histories, companies like Cashfree can provide more accurate nudges to consumers than generic third-party tools. This creates a competitive moat that is increasingly difficult for smaller players to replicate without similar scale.
Market participants will now monitor the adoption rates of these agents among small and medium enterprises. The transition to outcome-based pricing will serve as a critical test of whether merchants perceive the automated recovery as a direct driver of profitability. Future milestones will likely involve the expansion of these agents into more complex financial products and cross-border payment scenarios.
At Global Business & Economics Journal, we contribute to the future of business and technology with news, innovative research, expert insights, and thought-provoking analysis.
