Unit21 raises $13 million to turn AI into a fraud-fighting machine

Unit21, a startup providing a no-code service that monitors for fraudulent activity online, has raised $13 million in funding. The San Francisco-based company says the money will help grow 26-employee Unit21’s product and distribution-focused management teams, bolster sales and marketing efforts, and launch in new industries.

Of the 3.2 million identity theft and fraud claims received by the U.S. Federal Trade Commission in 2019, 1.7 million were fraud-related. Consumers reported losing more than $1.9 billion related to fraud complaints, an increase of $293 million from 2018. That’s in line with a trend identified in a recent Javelin Strategy & Research study: Identity fraud victims’ out-of-pocket fraud costs more than doubled from 2016 to 2018 to $1.7 billion.

Unit21, which was founded in 2018 by former Affirm product manager Trisha Kothari, aims to protect businesses against suspicious activity through APIs and dashboards for detecting fraud, money laundering, and other cross-industry risks. The platform offers over 100 out-of-the-box machine learning and threshold-based rules driven by a case management system, and it provides reporting and experimenting tools for big data analytics. Moreover, Unit21 integrates with products from Jumio, Trulio, and Dow Jones, affording access to things like identity verification and document checking.

Unit21’s case management system creates an immutable and auditable trail for all data and actions within a system. Users can customize workflows and actions to tailor the platform to their needs, exploring their own alerts and cases while spotlighting data in real time. Unit21 leverages machine learning to monitor transactions, applying different monitoring logic and thresholds to different segments of customers. And the platform’s operations management component enables clients to manage analyst workforces along dimensions like geography, product line, customer segment, and suspicious activity type.

“Unit21 allows non-technical teams to write intelligent rules that leverage historical data, like transaction history,” Kothari, who cofounded Unit21 with Clarence Chio, told VentureBeat via email. “Standard transaction monitoring for anti-money laundering can be very rudimentary. Black box machine learning won’t work in the compliance space since it is crucial to demonstrate to regulators that there is no bias … Unit21’s key focus is explainability, in the sense that risk and compliance teams have complete control over thresholds and parameters used to detect suspicious activity and hence can reap benefits of more intelligent rules without losing control.”

The fraud prevention space is filled with rivals making similar claims, including Vesta, Bleckwen, Arkose Labs, and Buguroo. But Kothari says that Unit21 has grown substantially this year, with a 400% uptick in business within the past 12 months and new customers like Intuit, Coinbase, Gusto, and Line.

Gradient Ventures, Google’s AI venture fund, contributed to Unit21’s tranche with participation from Core VC, South Park Commons, VMWare founder Diane Greene, Plaid founder William Hockey, Chime cofounders Chris Britt and Ryan King, Shape Security cofounder Sumit Agarwal, and former Venmo COO Michael Vaughan. It’s the company’s first priced fundraising round.