RBIH to build AI-driven Digital Payments Intelligence Platform; looks to hire senior leaders

Sahil Kini CEO RBIH

The Reserve Bank Innovation Hub (RBIH) - a subsidiary promoting inclusive access to financial services by leveraging technology - is building a new digital public infrastructure called the Digital Payments Intelligence Platform (DPIP) to use artificial intelligence (AI) and assign a real-time risk score to every digital transaction in India.

This scoring system, will help banks and payment service providers assess potentially fraudulent transactions before they are completed, creating a nationwide safety net for digital payments.

“The platform is still being designed and will be built over the next six months,” said Sahil Kini, CEO of RBIH, in an interaction with Moneycontrol.

“DPIP aims to provide a real-time risk score for every digital payment in India. That score will help banks decide what to do with each transaction,” he added.

How does DPIP work?

The platform will not share raw transaction data, instead, it will provide a risk signal that indicates if a payment is potentially suspicious.

“We are not giving out the data, only a signal of risk. The aim is to help institutions make faster, data-backed decisions without compromising privacy,” Kini clarified.

The initiative comes as India has witnessed a surge in UPI and digital payment frauds, phishing scams and mule account activities.

The DPIP is designed to act as a national intelligence layer, pooling data and insights from banks, fintechs and telecom operators to identify high-risk behaviour in real time.

What are DPIP’s capabilities?

The DPIP is being developed as a tool for AI-driven fraud detection, using machine learning to spot suspicious patterns. The platform will offer real-time intelligence sharing across banks and payment players, along with a centralized fraud registry which will track repeat offenders, mule accounts, and fraudulent phone numbers.

DPIP also includes a pre-transaction alerts enabling banks to act before funds are lost.

DPIP’s rollout will happen in phases, starting with a negative registry of known fraudulent accounts, expanding into advanced predictive models that anticipate fraud patterns.

What is Mule Hunter by RBIH?

Alongside DPIP, RBIH is also developing Mule Hunter, an AI model designed to detect mule accounts, or accounts used to launder stolen or illicit funds.

“The model has performed very well, with multiple banks adopting it,” Kini said. “Its accuracy is roughly 90%, and it has helped detect a much larger number of suspicious accounts. The response so far has been phenomenal.”

Is RBIH hiring for AI capabilities?

Both DPIP and Mule Hunter are being developed entirely in-house by RBIH, which is currently hiring engineers and senior leaders to scale its AI and data science capabilities.

“We are building them in-house. We are hiring engineers and senior leaders for this. You might have seen our LinkedIn posts for senior management hiring,” Kini added.