Mumbai-based fintech startup Data Sutram has raised $9 million in a Series A funding round led by global investors B Capital and Lightspeed. The fresh capital will be deployed to scale product development, expand the team, and pursue international growth, said Rajit Bhattacharya, Founder and CEO of Data Sutram, in an exclusive interview with CNBC-TV18.Currently focused on fraud detection during customer onboarding for banks and NBFCs, the company plans to deepen its capabilities by moving into transaction-level monitoring. “We’re currently solving for fraud at the onboarding level, but we eventually want to go into transaction-level monitoring, looking at payments and other aspects,” Bhattacharya explained.Data Sutram is also expanding its reach into sectors beyond traditional lending, such as insurance, crypto, and gaming — industries with complex digital money trails and heightened fraud risks. “Wherever there’s a money trail involved digitally, there’s a potential risk of fraud, and that’s where we plan to expand our product portfolio,” he said.
Operating on a pay-per-inquiry billing model, Data Sutram reported a 4x revenue growth last fiscal year (FY25), reaching a $3 million run rate. The company aims to replicate this growth and hit $10 to $12 million in the next 12 to 15 months.The startup’s technology is rooted in artificial intelligence and machine learning that analyse an individual’s digital and social behaviour to generate a proprietary trust score. Unlike traditional credit scores that rely on borrowing history, Data Sutram’s system looks beyond to detect subtle signs of fraud from millions of digital footprints.Bhattacharya highlighted the significance of building a homegrown fraud detection system, saying, “It was a big, big win for us, especially in a segment like financial services, for large banks and large lenders to come in and trust an indigenous, made-in-India proprietary tech, and to trust it for core decision-making to solve fraud.”He added that the company’s approach does not rely on predefined fraud patterns. “At its core, our technology assumed we didn’t know what fraud looked like. We built machine learning and AI-based models that were able to come up with our own score that captures what fraud could look like.”Watch the accompanying video for more
