Hello Clever protects every transaction through a layered fraud prevention system that combines rule-based detection, machine learning models, automated blocking, human compliance review, and post-payment screening. This page explains how the fraud monitoring system works, what signals it uses to flag suspicious activity, how flagged transactions are handled, and how Hello Clever secures real-time and gateway payments against fraud.

​
How fraud detection works

Hello Clever’s transaction monitoring applies both rule-based and machine learning (ML) detection in real time. Rules are tuned to catch fraud without interrupting legitimate customers, while ML models continuously adapt to evolving fraud patterns. Hello Clever’s fraud detection is further enhanced through a partnership with Flagright, a specialist in AML compliance and transaction monitoring.

​
Fraud signals monitored

Transactions that deviate significantly from a user’s historical spending patterns trigger an alert. The system accounts for gradual spending growth so legitimate increases do not create false positives.Example: A user who typically spends 50suddenlyattemptsa50 suddenly attempts a 2,000 transaction.
Purchases originating from unexpected or high-risk locations are flagged. The system checks for travel signals (such as recent airline purchases) before escalating, to avoid blocking genuine travellers.Example: A customer who normally transacts in Sydney makes a purchase from Moscow.
A sudden shift in device, browser, or operating system can signal account takeover. Known device upgrades and previously used devices are recognised and treated as lower risk.Example: A user who always transacts from an iPhone suddenly initiates a payment from an unrecognised device.
A high number of transactions in a short window may indicate automated scripts or malicious activity. The system adjusts sensitivity during high-frequency shopping periods like major sales events.Example: Five transactions within under a minute from a new device.
A sudden shift in spending category or value outside a customer’s established patterns is flagged. Recent salary deposits or seasonal spending spikes are factored in to reduce false positives.Example: A customer with steady grocery and utility spend suddenly makes multiple high-value luxury purchases.
Repeated failed attempts in quick succession may indicate card testing or an account takeover attempt. The system distinguishes between network retries and genuinely suspicious failure patterns.Example: Multiple failed transactions from different devices or locations in quick succession.

​
How flagged transactions are handled

When the fraud detection system identifies a suspicious transaction, one of the following responses is triggered:

Automated blocking

Transactions matching known high-risk patterns (such as flagged geolocations or confirmed fraud signatures) are blocked automatically in real time before processing completes.

Compliance review

Transactions that show unusual patterns without meeting the threshold for automatic blocking are escalated to Hello Clever’s compliance team for manual investigation.

Customer verification

Where further verification is required, customers may be contacted to confirm their identity before the transaction is released or permanently declined.

Post-payment screening

Completed transactions are retrospectively reviewed to identify patterns (such as coordinated fraud rings) that only become apparent over time.

​
Post-payment screening

In addition to real-time monitoring, Hello Clever conducts regular post-payment screening to detect fraud that is not immediately apparent at transaction time.
  • Ongoing behavioural analysis: completed transactions are reviewed for patterns that may collectively indicate risk, even when individual transactions appear legitimate.
  • Retrospective review of high-risk transactions: previously flagged transactions are periodically re-examined to identify links to fraud rings or coordinated attacks.
  • Continuous ML model improvement: post-payment data feeds back into machine learning models, refining detection accuracy for evolving fraud tactics.

​
Flagright partnership

Hello Clever’s fraud detection capabilities are extended through its partnership with Flagright.
  • Enhanced pattern recognition: Flagright identifies subtle correlations across transaction data that may escape standard rule-based systems.
  • Real-time threat intelligence: Flagright provides up-to-date intelligence on emerging fraud trends, allowing Hello Clever to adjust detection rules proactively.
  • Customisable rule sets: the partnership enables Hello Clever to refine monitoring rules in response to the latest fraud tactics without delays.

​
Payment gateway security

Hello Clever’s payment gateway adds further layers of protection beyond transaction-level fraud detection.
Customers sign in using a one-time password (OTP) sent to their registered phone number, eliminating the risks associated with static passwords.
Cloudflare secures Hello Clever’s platform against DDoS attacks, SQL injection, and other common web threats, providing real-time blocking at the network edge.
Hello Clever’s infrastructure is continuously monitored by Azure Security Centre, which uses advanced threat detection and automated alerts to identify risks before they impact operations.
Hello Clever undergoes routine penetration testing to proactively identify and remediate vulnerabilities. Contact Hello Clever’s support team to request a penetration test report.

​
Real-time payment security

Real-time payments (such as PayID-based account-to-account transfers in Australia) offer inherent security advantages over card payments that Hello Clever leverages across its platform.

Direct bank authentication

Payments are authenticated directly within the customer’s bank using multi-factor authentication (MFA), with no card data transmitted or stored.

Reduced fraud surface

Fewer intermediaries means fewer potential access points for fraud. No card numbers or CVVs are involved in account-to-account transfers.

No chargebacks

Real-time payments are final and irrevocable once confirmed. This eliminates chargeback risk and preserves your revenue.

Simplified PCI compliance

Without card data handling, your business faces reduced PCI-DSS obligations, lowering operational complexity and cost.
Hello Clever applies end-to-end encryption and continuous fraud monitoring to all real-time payments, with risk-based scoring on every transaction and payee verification before funds are transferred.
Offering real-time payment methods such as PayID alongside card payments gives your customers a low-friction, high-security option that also reduces your exposure to card fraud and chargebacks.