Introduction
Sardine is an agentic AI risk platform that helps banks, fintechs, and other financial institutions detect fraud, screen for money laundering, and support credit underwriting from a single system. It was founded in 2020 by Soups Ranjan, Aditya Goel, and Zahid Shaikh, three former colleagues who had built in-house fraud and compliance infrastructure together at Revolut before leaving to start the company.
Sardine’s name is a deliberate play on Suspicious Activity Reports, the filings financial institutions submit when they flag potential fraud or money laundering. For readers of Brel’s AI in finance coverage, Sardine is a useful example of the shift from AI models that merely score risk toward AI agents that carry out much of the investigation and compliance workflow themselves.
What the company does
Sardine’s platform unifies device intelligence, real-time transaction monitoring, onboarding and identity checks, anti-money-laundering screening, and account-takeover protection into a single risk system, with AI agents that the company says can execute the full investigation lifecycle — from initial detection through case review to regulatory filing — rather than only surfacing alerts for a human analyst to work through manually.
Who it serves
Sardine’s own materials describe serving more than 300 enterprises across roughly 70 countries, including companies such as FIS, GoDaddy, Deel, Checkout.com, Brex, and, at various points, X (formerly Twitter). The company has also built relationships with financial institutions directly, including a 2026 partnership expansion with National Bank of Canada across its retail, commercial, and wealth operations.
Company background
Soups Ranjan, who holds a PhD in electrical and computer engineering from Rice University, previously led risk functions at Coinbase and Revolut before co-founding Sardine with Aditya Goel and Zahid Shaikh, who worked alongside him building fraud and compliance infrastructure at Revolut; the three left in April 2020 to start the company in San Francisco. Sardine’s early customer base was concentrated in crypto, including companies such as FTX and MoonPay, before it expanded into broader banking and fintech risk management. The company raised a $70 million Series C round in February 2025 led by Activant Capital, with participation from Andreessen Horowitz, Nyca Partners, Google Ventures, and other investors, bringing total funding to $145 million at a reported $660 million valuation. In May 2026, Sardine raised a $25 million Series C extension led by National Bank of Canada — which had already invested through its NAventures arm — bringing total funding to $170 million.
Product and AI capabilities
Sardine applies AI models to device and behavioral signals — the company says it has profiled more than 2.2 billion devices — to score risk in real time during onboarding, transactions, and account activity, while positioning its more recent AI agents as tools that automate the repetitive work of clearing transaction alerts, verifying onboarding cases, and investigating fraud rings, rather than simply summarizing them for a human. In March 2026, CEO Soups Ranjan described using Sardine’s agentic tools to help expose a 150,000-account fraud ring in 11 minutes, an example the company has used to illustrate how automation can compress investigation timelines that previously took risk teams far longer.
Key developments
Sardine was founded in 2020 by Soups Ranjan, Aditya Goel, and Zahid Shaikh, following their work together on fraud and compliance systems at Revolut. The company raised a $70 million Series C round in February 2025 led by Activant Capital. In May 2026, it raised a $25 million Series C extension led by National Bank of Canada, bringing total funding to $170 million, alongside a multi-year partnership expansion covering the bank’s retail, commercial, and wealth operations.
Why it matters
Sardine is a useful marker of how fraud-prevention vendors are responding to rising alert volumes — which the company has said have grown sharply as fraud itself becomes more automated and AI-assisted — by moving toward AI agents that execute investigations rather than only flag them. That shift raises its own questions about auditability and human oversight in regulated compliance workflows, which is part of why Sardine’s public materials emphasize that its models are designed to be explainable and to work alongside, not replace, human risk analysts. Sardine’s origins in crypto-focused fraud prevention, before expanding into banks and traditional fintechs, is also a reminder that some of the more aggressive early adopters of agentic risk tooling have been companies operating in less-regulated or newer markets, where there was less incumbent tooling to displace.
Sector context
Sardine sits within Brel’s AI in finance coverage alongside other fraud and compliance platforms such as Feedzai and identity-verification vendors such as Socure, though its emphasis on agentic automation of the full investigation-to-filing workflow reflects a newer generation of risk-technology product design. Its overlapping coverage of fraud, compliance, and credit-underwriting signals also places it in a similar cross-functional category to Unit21, another vendor consolidating previously separate risk functions into one AI-driven platform.
Sources and references
This profile draws on Sardine’s own “About Us” page and its official funding announcements covering its 2025 Series C round and 2026 Series C extension.
- Sardine — “About Us” company page (sardine.ai)
- fintech.global — Sardine Series C funding coverage (2025)
- fintech.global — Sardine Series C extension coverage (2026)