.png){kind=logo}
As digital payments scale rapidly, so do the risks of fraud across the ecosystem. In this episode of the Analytics Insight Podcast, host Priya Diyalani speaks with Deepak Chandnagar, CEO and Co-founder of NPST, about the rising threat of payment fraud in the age of digital commerce. The discussion explores how AI-powered risk intelligence platforms are transforming fraud detection, merchant monitoring, and enterprise decision-making
A: NPST has been operating for over a decade, initially focused on building digital payment infrastructure for underserved regions in India. Over time, we expanded into core banking and payment technologies, including UPI, IMPS, and mobile banking. Today, we process around 7–8% of India’s UPI volume and work with over 20 banks and 100+ clients. Our focus is on building scalable, compliant payment ecosystems. Recently, we’ve expanded into AI-driven RegTech to address fraud risks across digital payments.
A: Risk is no longer limited to fraudulent transactions, it spans the entire merchant lifecycle. Payment ecosystems are highly interconnected, and vulnerabilities can emerge at onboarding, transaction processing, or post-transaction stages. Fraud prevention must be embedded into the digital growth strategy rather than treated as a separate compliance layer. Today, monitoring merchant behavior, transaction patterns, and anomalies in real time is critical, especially at scale where manual oversight is impossible.
A: RIDP is an AI-powered platform designed to establish trust across merchants and transactions. It assigns a “trust score” by analyzing multiple parameters, similar to underwriting in lending. We evaluate merchant authenticity, business behavior, transaction limits, and external data sources like GST or web presence.
Once onboarded, every transaction is monitored using AI trained on hundreds of millions of transactions. The system identifies behavioral anomalies, for example, unusual transaction spikes at odd hours and flags risks in real time. Over time, the platform evolves by learning patterns and generating its own rules, making it more adaptive than traditional systems.
A: Rule-based systems are static, while fraud is dynamic. Fraudsters eventually learn system rules and find ways to bypass them. Moreover, traditional systems primarily focus on pre-transaction checks. The real challenge lies in post-transaction fraud, where suspicious activities surface only after execution. AI-driven systems, on the other hand, continuously analyze behavior, adapt to new fraud patterns, and provide predictive insights rather than reactive responses.
A: We’re seeing increasing fraud in post-transaction scenarios, KYC manipulation, and mule account networks. Fraudsters are becoming more sophisticated, exploiting gaps in onboarding and transaction monitoring. For instance, a merchant may appear legitimate during onboarding but later misuse payment channels. Detecting such patterns requires continuous monitoring, behavioral analysis, and AI-driven anomaly detection across the entire lifecycle.
A: AI allows us to move from reactive fraud detection to proactive intelligence. By analyzing transaction behavior, merchant activity, and contextual data, AI can identify patterns that humans cannot detect at scale. It can also predict potential fraud risks, highlighting areas where anomalies may occur in the future. This enables fintechs and banks to act before fraud escalates, reducing financial losses and improving operational efficiency.
A: The biggest challenge is mindset. Many organizations still rely on traditional, regulation-driven systems and hesitate to adopt new approaches. Accepting AI as a core requirement, not an optional add-on, is critical.
Second is understanding AI’s potential. Many organizations lack clarity on how AI can enhance fraud detection, from merchant onboarding to real-time monitoring.
Finally, there’s the need for a forward-looking approach. Organizations must shift from controlling known fraud risks to predicting future threats. This requires investment in data, technology, and a willingness to evolve beyond legacy systems.
The future lies in AI-driven, multi-layered risk intelligence. Fraud management will become a continuous, real-time process integrated across the entire payment lifecycle. Organizations that combine AI capabilities with domain expertise will be better equipped to scale securely while maintaining trust in the digital payments ecosystem.
Listen to the podcast to know more about this.
