Sardine
Sardine Technical Briefing
Published July 13, 2021
Supported by IowaEDA
Overview
Soups Ranjan, CEO and co-founder of Sardine, walks through the Sardine fraud and compliance API — a single endpoint for fintechs, crypto exchanges, challenger banks, and NFT platforms to prevent fraud, run KYC, monitor crypto transactions, and stay on top of AML and sanctions (PEP/SDN) screening. The demo shows how Sardine catches fraudsters running mobile emulators like BlueStacks, disposable phone numbers, and stolen dark-web identities — combining behavioral biometrics, device intelligence, and telco/social-graph enrichment with live Chainalysis and Coinbase Analytics calls. A no-code rule editor lets ops analysts chain dozens of fraud and AML typologies without writing code, and adverse-media signals cut AML false positives that plague traditional transaction-monitoring rules.
0:00 Introduction to Sardine
Soups Ranjan introduces Sardine — the fraud and compliance API-as-a-service for fintechs, crypto on-ramps, exchanges, NFT marketplaces, and cross-border commerce.
0:30 Stripe for high-risk merchants
How Sardine positions itself as a single API for fraud prevention and AML compliance across the hardest merchant categories, with founding-team experience at Coinbase and Revolut.
1:30 The fraudster's mindset
Live walk-through from the attacker's side — VPNs, mobile emulators like BlueStacks, stolen dark-web identities, disposable emails, and prepaid phone numbers used to open fintech and crypto accounts.
3:00 Behavioral biometrics at onboarding
How Sardine captures typing, tapping, swiping, mouse movement and context switches — so long-term-memory fields like name, address, and SSN are scored differently for real customers vs. fraudsters copy-pasting dark-web data.
4:30 ML fraud scoring with Chainalysis and Coinbase Analytics
Supervised and unsupervised ML models score each customer across a consortium of tens of millions of fintech and crypto consumers, with live calls to Chainalysis and Coinbase Analytics for crypto withdrawal risk.
5:30 KYC: sanctions, PEP and SDN screening
Single-API KYC covering SSN verification, sanctions screening, politically-exposed-person checks, and suspected designated-nationals matching — all returned alongside the fraud score.
7:30 Identity correlation across telco, social, and bureau
How Sardine correlates telco registration, social-media footprint, bureau identity, and document-KYC to flag mismatches — the dashboard surfaces all of them per user for analyst review.
9:00 No-code rule editor for fraud and AML typologies
Dozens of built-in fraud and AML rule typologies (withdrawal to multiple high-risk crypto addresses, name mismatch vs bank account, high-risk carrier, low social graph) that ops analysts can chain without writing code.
10:30 Cutting AML false positives with adverse-media signals
Why traditional AML rules (just transaction amount and count) produce too many false positives, and how Sardine combines adverse-media, crypto address risk, and device/behavior signals for a holistic AML picture.
Presented by Soups Ranjan and Aditya Goel — Sardine · website
Topics: Compliance & Regulation, Identity & KYC, Payments
