Catch money mules before the money moves.
Drop in any transaction export: Excel, CSV or a core-banking dump. RiskGPT works out the format by itself, runs 12 mule-detection checks and flags suspicious accounts in seconds. All of it happens inside your browser. Not one row leaves your machine.
Free for 3 files · No sign-up · Zero upload
Live alert feed
43/43
mules caught
0 KB
uploaded
<100ms
5,000 txns
The problem
The global fraud economy runs on mule accounts.
Authorised push-payment scams, digital-arrest calls, fake trading apps and phishing all need one thing: a bank account to receive and move stolen money. Students, gig workers and dormant accounts are rented out as mules. Legacy AML runs overnight — by the time an alert fires, the money is gone and the bank faces freeze orders, customer complaints and regulator questions.
$1 trillion+
estimated lost to scams worldwide each year (GASA)
< 10 min
for stolen money to hop through 3–5 mule accounts on instant rails — UPI, FedNow, Faster Payments, SEPA Instant
₹22,800 cr+
lost to cyber fraud reported in India alone in 2024 (I4C)
Privacy by architecture
Your data never leaves the browser. Not even once.
Bank statements contain account numbers, UPI IDs and customer PII. Uploading them to a SaaS server is a compliance risk. RiskGPT downloads the engine to your device and runs everything locally — parsing, detection, scoring, case notes and exports.
How zero-upload worksRuns on your CPU
Processed in browser memory only.
Works offline
Load the page, turn off Wi-Fi, still scans.
Zero upload
Check DevTools → Network: no rows sent.
GDPR, DPDP & bank-secrecy friendly
Nothing stored on our servers, ever.
Regulatory fit
Aligned with regulators worldwide.
Global · FATF Recommendations
Risk-based transaction monitoring and suspicious-activity reporting — the baseline every AML regime builds on.
India · RBI Fraud Risk Directions (2024)
Banks must run early-warning signals and proactive transaction monitoring to detect fraud and mule accounts.
India · PMLA, FIU-IND STR & DPDP Act
Ongoing due diligence, STR filing to FIU-IND, and minimal movement of personal data.
US · FinCEN / Bank Secrecy Act
Monitor for suspicious activity and file SARs; mule networks are a FinCEN advisory priority.
UK · FCA, POCA & UK GDPR
Systems to detect money laundering, SARs to the NCA, and APP-fraud reimbursement pressure.
EU · AMLD6 / AMLA & GDPR
Harmonised AML supervision under AMLA with strict limits on moving personal data.
Privacy by design
GDPR, DPDP and bank-secrecy friendly — RiskGPT processes everything on your device.
RiskGPT supports your compliance programme; it doesn't replace your bank's own policies or legal advice. STR/SAR drafts are reviewed and filed by your Principal Officer or MLRO.
Technology
Forensic-grade engine. Zero infrastructure.
In-browser forensic engine
TypeScript detection core runs 12 typologies over lakhs of rows in milliseconds — on your laptop's CPU.
Graph fund-flow analysis
Traces remitter → mule → beneficiary chains, circular round-trips and fan-in / fan-out rings.
Device, IP & geo signals
Shared-device rings, multi-city ATM cash-outs and KYC mismatches from your account master.
Explainable scoring
Every score shows the rule, the evidence and the exact transactions. No black box, audit-ready.
Connect directly to your core banking system
Enterprise on-prem engine plugs into Finacle, Flexcube, BaNCS and UPI switch feeds for near real-time mule alerts — inside your own data centre.
12 mule typologies, built for global banking
Threshold rules miss mules. These look at speed, spread, devices, geography and behaviour change.
Rapid pass-through
Money received is sent out again (90%+) within an hour, many times.
Fan-in collector
15+ different remitters pay one account within 48 hours.
Fan-out dispersal
One account pays 10+ different beneficiaries within 48 hours.
Cash structuring
Repeated cash deposits just under the local reporting threshold, often across cities.
Nil retention
Outflow almost equals inflow — nothing stays in the account.
Dormant reactivation
Quiet for 6+ months, then a test credit and large money movement.
Layering chain
Funds hop between own-bank accounts within hours with tiny deductions.
Circular round-trip
The same amount goes around a loop of accounts and comes back.
New account velocity
Newly opened or min-KYC account with turnover far above declared income.
Shared device ring
Several different customers using the same phone, device or mobile number.
Multi-city ATM cash-out
ATM withdrawals in 3+ cities within a few hours — physically impossible for one person.
Profile change
Normal history, then a new device, new beneficiaries and a jump in value.
Who it's for
Banks
Fraud & AML teams triaging scam complaints — 1930/NCRP, APP fraud, Action Fraud.
Fintechs & PSPs
Screen merchant, wallet and instant-payment flows.
NBFCs & lenders
Check disbursal recipients before money leaves.
Forensic auditors
Client NDAs forbid uploads. Analyse locally.
FAQ
Find the mules hiding in your data — in under a minute.
Scan my fileFree for 3 files or 100,000 rows · nothing uploaded