Frequently Asked Questions
Everything you need to know about Fraudalysis on-chain fraud detection, how it works, and how it can protect your platform.
What is Fraudalysis and how does it detect blockchain fraud?
Fraudalysis is a real-time on-chain surveillance platform that monitors blockchain activity continuously using machine learning. It detects suspicious patterns — rapid senders, large transfers, wash trading, and money-laundering behaviour — by analysing transactions block by block as they occur on the network, without relying on static blacklists.
What types of on-chain fraud does Fraudalysis detect?
Fraudalysis detects a wide range of on-chain threats including rug pulls, rapid sender patterns (many transactions from the same address in a short window), large anomalous transfers, new agent registrations on identity contracts, wash trading, drainer kit deployments, and money-laundering patterns that move funds through mixers or intermediary wallets.
How does real-time blockchain monitoring work?
Real-time monitoring works by polling the blockchain network every few seconds for new transactions and blocks. Each transaction is analysed against behavioural ML models — looking at sender frequency, transfer amounts, contract interactions, and historical patterns — to flag anomalous activity within milliseconds of it appearing on chain.
Can Fraudalysis prevent rug pulls before they happen?
Fraudalysis cannot prevent a rug pull directly (it does not control smart contracts), but it can detect the warning signs early — unusual liquidity movements, rapid token transfers, or sudden ownership changes — and alert platform operators in real time so they can take protective action before investor funds are drained.
How does Fraudalysis detect rapid senders and suspicious activity?
Rapid sender detection monitors the number of transactions originating from a single address within a defined time window. If an address sends 5 or more transactions in a short period — especially to different recipients — it is flagged as a rapid sender. This pattern is commonly associated with dusting attacks, sybil activity, and money-laundering preparation.
Is Fraudalysis compatible with the MultiversX blockchain?
Yes, Fraudalysis currently monitors the MultiversX devnet (MX-8004) as a prototype deployment, monitoring transaction activity, identity contract registrations, and rapid sender patterns. Mainnet support for MultiversX and other networks is in active development. Contact us for specific network integration timelines.
How does machine learning detect fraud patterns on-chain?
Fraudalysis uses behavioural ML models trained on thousands of known fraud scenarios — rug pulls, sandwich attacks, wash trading, and money-laundering chains. Instead of matching static blacklists (which become stale), the models learn the underlying behaviour patterns, allowing them to detect novel attacks they have never seen before.
What is a rapid sender alert and why is it flagged?
A rapid sender alert triggers when a blockchain address initiates an unusually high number of transactions within a short time frame (e.g., 5+ transactions in under a minute). This behaviour is characteristic of automated bot networks, sybil attackers, and money-laundering operations that rapidly move funds between wallets to obscure the trail.
How does wallet address screening work?
Address screening evaluates a wallet's activity by analysing its on-chain behaviour — who it has interacted with, its transaction frequency profile, and any involvement in previously flagged activities. Activity history for an address can be queried via the public API, and automated risk scoring is live at GET /api/v1/risk/:prefix — returns a behavioural score 0–100 with tier labels and detailed reasons.
What is the difference between blacklists and behavioural ML detection?
Blacklists are static lists of known bad addresses — they only catch repeat offenders and miss new ones. Behavioural ML detection analyses how addresses behave: transaction frequency, value patterns, recipient diversity, and contract interactions. This means Fraudalysis can flag a brand-new address as suspicious based on its behaviour alone, without ever having seen it before.
Can I integrate Fraudalysis with my platform via API?
Yes — Fraudalysis is API-first. A public data feed and REST API endpoints for alert data and transaction analysis are live at /api/v1, letting you integrate real-time fraud detection into your onboarding flow, transaction approval pipeline, or compliance dashboard. Contact us to discuss integration.
How accurate is Fraudalysis's detection and how are false positives handled?
Fraudalysis's ML models are trained on labelled fraud datasets and continuously improved through feedback loops. False positives are minimised through multi-factor analysis (combining behavioural, transactional, and temporal signals). Every alert is logged with full context so operators can review, whitelist, or escalate as needed.
What happens when a suspicious transaction is detected?
When a suspicious pattern is detected, Fraudalysis logs the finding with a severity level (WARN, ERROR, or INFO) and the relevant transaction details. A real-time alert is sent via the configured channel (currently Telegram with batched summaries). The finding is also recorded in the public results dashboard for review and analysis.
How does Fraudalysis help with AML compliance?
Fraudalysis supports AML (Anti-Money Laundering) compliance programs by providing continuous transaction monitoring, rapid sender detection, and large transfer flagging. These capabilities give compliance teams a documented evidence trail to support their own suspicious activity reporting and ongoing due diligence. Fraudalysis is a monitoring and analytics tool — it does not provide legal advice or file regulatory reports; platforms remain responsible for meeting their own obligations.
How often does the fraud monitor scan the blockchain?
The fraud monitor scans the blockchain approximately every 6 seconds (aligned with block production intervals). Each scan checks for new transactions, large transfers, rapid sender patterns, and new agent registrations. The system operates 24/7/365 with all findings logged and displayed on the public results dashboard.
Still have questions? Contact us and we will be happy to help.