Custom Transaction Monitoring System for AML Compliance

You launch a crypto exchange on Base or Ethereum L2. Deposits grow, but the compliance team drowns in manual checks. One missed structuring pattern—and the regulator issues a fine. Every transaction is a potential risk: without automated monitoring you miss up to 15% of suspicious operations. A transaction monitoring (TM) system analyzes each operation in real time, flagging anomalies by velocity, structuring, and other patterns. We have built dozens of such systems for exchanges and DeFi projects, guaranteeing compliance with FATF and local regulators.

What problems does transaction monitoring solve?

Transaction monitoring is not a blacklist of addresses. It is continuous analysis: velocity, structuring, round-trip transfers, geographic anomalies. Example: a client transfers $450k in 24 hours when their average daily volume is $2k—a 225x ratio. The rule-based engine immediately flags a MEDIUM alert; the ML check confirms the anomaly at 98.7%—a full freeze is triggered.

Typical patterns we handle:

• Structuring: several transactions of $9,500 over 3 days to circumvent the $10k reporting threshold.
• Velocity: 15 transactions in one hour from different IPs—sign of a bot.
• Round-trip: deposit $50k, withdraw to the same addresses minus fee after 6 hours.

How we build the monitoring system—a client case

One of our projects was an exchange on Arbitrum with 200k active users. Initially they used a third-party API for address checks—12% of suspicious transactions slipped through. We deployed a hybrid architecture:

The rule engine contains 14 deterministic rules (TM-001–TM-014). The ML module is retrained weekly on historical data. Results: zero false negatives over 8 months, detection time 86 ms.

Example Structuring Rule (TM-001)

const STRUCTURING_RULE: MonitoringRule = {
  id: "TM-001",
  name: "Structuring Detection",
  category: "structuring",
  alertLevel: AlertLevel.HIGH,
  action: AlertAction.FREEZE_AND_REVIEW,
  
  async evaluate(ctx: TransactionContext): Promise<RuleResult> {
    const REPORTING_THRESHOLD = 10000;
    
    // Find transactions just below threshold in the last 3 days
    const nearThreshold = ctx.history30d.filter(t => 
      t.usdAmount >= REPORTING_THRESHOLD * 0.7 &&
      t.usdAmount < REPORTING_THRESHOLD &&
      Date.now() - t.timestamp < 3 * 86400000
    );
    
    const currentNearThreshold = ctx.transaction.usdAmount >= REPORTING_THRESHOLD * 0.7 &&
                                 ctx.transaction.usdAmount < REPORTING_THRESHOLD;
    
    if (currentNearThreshold && nearThreshold.length >= 2) {
      return {
        triggered: true,
        alertLevel: AlertLevel.HIGH,
        details: `${nearThreshold.length + 1} transactions just below $${REPORTING_THRESHOLD}`,
        evidence: nearThreshold.map(t => t.id),
      };
    }
    
    return { triggered: false };
  },
};

ML-based Anomaly Detection

from sklearn.ensemble import IsolationForest
import numpy as np

class TransactionAnomalyDetector:
    def __init__(self):
        self.model = IsolationForest(contamination=0.01, random_state=42)
    
    def extract_features(self, transaction, user_history):
        return [
            transaction['usd_amount'],
            transaction['usd_amount'] / (user_history['avg_30d'] + 1),
            len(user_history['transactions_24h']),
            transaction['hour_of_day'],
            transaction['day_of_week'],
            user_history['unique_counterparties_7d'],
            transaction['aml_risk_score'],
        ]
    
    def predict(self, features) -> float:
        # Returns: -1 anomaly, 1 normal
        # Transform to probability
        score = self.model.score_samples([features])[0]
        return (score + 0.5) * 2  # normalize to [0, 1]

Why we use rule-based + ML

Rule-based is faster to interpret; ML catches what isn't explicitly written. In practice: rules cover 80% of known schemes, ML adds another 15%, the rest are false positives that require an operator to review. A pure rule-based system yields about 2% false positives; our hybrid gets 0.5% with the same recall.

Rule-based vs ML comparison

Alert Management and SAR (Suspicious Activity Report)

class AlertManager {
  async createAlert(tx: Transaction, rules: RuleResult[], action: AlertAction): Promise<Alert> {
    const alert = await this.db.createAlert({
      transactionId: tx.id,
      userId: tx.userId,
      triggeredRules: rules.map(r => r.ruleId),
      maxAlertLevel: Math.max(...rules.map(r => r.alertLevel)),
      action,
      status: AlertStatus.OPEN,
      assignedTo: await this.autoAssignCompliance(),
      dueDate: this.calculateDueDate(action),
    });
    
    if (action === AlertAction.FREEZE_AND_REVIEW) {
      await this.freezeUserAccount(tx.userId, alert.id);
    }
    
    await this.notifyComplianceTeam(alert);
    return alert;
  }
  
  async resolveSARAlert(alertId: string, sarDecision: SARDecision): Promise<void> {
    if (sarDecision.submitSAR) {
      await this.sarService.createAndSubmit({
        alertId,
        suspiciousActivity: sarDecision.description,
        supportingTransactions: sarDecision.transactions,
      });
    }
    
    await this.db.updateAlert(alertId, {
      status: sarDecision.submitSAR ? AlertStatus.SAR_SUBMITTED : AlertStatus.CLOSED,
      resolution: sarDecision.resolution,
      resolvedAt: new Date(),
    });
  }
}

Development process

1. Audit current compliance processes and transaction flows.
2. Design rules and ML models for your jurisdiction.
3. Implement rule engine and integrate with blockchain (RPC, mempool).
4. Test on historical data—validate coverage of at least 90%.
5. Deploy and train the team.

What’s included

• Rule engine with 14+ preconfigured rules (structuring, velocity, round-trip, geographic).
• ML module based on Isolation Forest with weekly retraining.
• Alert Manager with automatic SAR creation.
• Dashboard for the compliance team.
• API for integration with any platform.
• Test documentation and team training.

Estimated timeline

From 2 to 3 months—from audit to production. Urgent integration with basic rules—from 3 weeks. Pinpoint your case—we’ll evaluate within 2 days.

We have developed AML systems for 5 exchanges and 12 DeFi projects. Our experience in Ethereum and Solana smart contract formal verification allows us to integrate monitoring at the chain level. Contact us to discuss your project and get a demo.

Comparison of approaches: Rule-based detects known patterns (structuring, velocity) in 100% of cases; ML finds 30% of new attacks not covered by rules. Together—95% coverage of suspicious schemes with 0.3% false positives.

Why does your project risk without blockchain compliance services?

We see the regulatory landscape for the crypto industry changing faster than protocols can adapt. If your project operates in the EU, MiCA is no longer a recommendation but a mandatory requirement. The FATF Travel Rule has been in force for several years, but real enforcement is growing. Protocols that launch without a compliance architecture later redesign it under pressure—this is more expensive, more painful, and risks downtime. Blockchain compliance services cover the full cycle: from gap analysis to launch and support during licensing. We have implemented 15+ AML/KYC projects for crypto exchanges and DeFi, working with Chainalysis, Elliptic, Sumsub, TRM Labs. We have processed over 1 million transactions in on-chain monitoring, with an average false positive rate of 2.3% for AML screening.

Why is the Travel Rule a technical, not a legal challenge?

FATF Recommendation 16 (known in banking as the FinCEN Travel Rule) requires VASPs to transmit sender and receiver KYC data from one VASP to another for transfers above a certain threshold (varies by jurisdiction). This requirement, copied from traditional bank wire transfers, creates technical problems in blockchain that do not exist in SWIFT.

The first problem is determining VASP-to-VASP. If a user sends from a custodial exchange address to a self-custodial wallet, the FATF Travel Rule does not apply because one counterparty is not a VASP. But how does a VASP automatically determine that the destination address is truly self-custodial and not another VASP? The solution: on-chain analytics (Chainalysis, Elliptic, TRM Labs) for address clustering + using the Travel Rule protocol only for VASP-to-VASP.

The second problem is interoperability between VASPs. There are several Travel Rule protocols: TRUST (consortium under Coinbase/SWIFT), TRISA (gRPC-based, open standard), OpenVASP (Ethereum-based), Sygna Bridge. They are not interoperable. Most major exchanges support several simultaneously. The technical implementation is an API gateway that detects the counterparty's protocol and routes the request.

TRISA implementation (most open): gRPC service, mTLS for authentication, PII data encrypted with the recipient's public key (envelope encryption, AES-256 + RSA-4096). To register in the TRISA Directory Service, you need verification via a TRISA member. The code is an open SDK in Go and Python.

Specific pain point: timing. Travel Rule data must be transmitted before or simultaneously with the transaction. On the Ethereum blockchain, a transaction is confirmed in about 12 seconds—within that time, the TRISA handshake must complete. If the counterparty does not respond, the transaction is blocked or delayed. The UI must explain this to the user, otherwise a flood of support tickets is guaranteed.

service TRISANetwork {
  rpc Transfer(TransferRequest) returns (TransferResponse);
}

message TransferRequest {
  string identity_payload = 1;  // encrypted PII packet
  string envelope_public_key = 2;
  string transaction_hash = 3;
}

How to choose a KYC/AML provider for a crypto project?

KYC providers for cryptocurrencies fall into several tiers:

Tier 1 (enterprise, regulatory grade): Jumio, Onfido, Sumsub, Veriff. Support 200+ countries, video verification, liveliness checks, AML screening via Refinitiv/Dow Jones. Integration via REST API + webhooks. Sumsub is popular in European crypto projects—good SDK documentation for mobile apps.

Tier 2 (DeFi-native, privacy-focused): Fractal ID, Synaps, Persona. Less regulatory overhead, faster integration, but less global coverage for high-risk jurisdictions.

On-chain KYC via credentials: Quadrata Passport, Civic, PolygonID—user verifies once, gets an on-chain credential, protocols verify it without repeated verification. Privacy-preserving via ZK. Not mainstream yet, but we are laying the groundwork in the architecture.

Architectural principle: KYC data is never stored on-chain. Personal data is stored with the provider or in your encrypted database; on-chain only a hash (commitment) or credential (if using VC/SBT approach). This ensures GDPR compliance: the right to erasure is achievable if data is off-chain.

Typical mistake: storing wallet-to-identity mapping in plaintext in PostgreSQL without row-level encryption. One SQL injection and the entire KYC database is compromised. Minimum: column encryption for PII fields (PGP or AES via pgcrypto), separate key management (AWS KMS, HashiCorp Vault), audit log for all PII access.

For AML screening, we use Chainalysis, Elliptic, or TRM Labs. Integration is asynchronous via webhook: results come in 1–5 seconds. Threshold-based blocking: HIGH risk — auto-block, MEDIUM — manual review. Hold period for suspicious transactions is 24–72 hours until manual review. Sanctions screening separately: OFAC SDN list updates several times a week; we use direct OFAC list integration (free) with custom address matching logic.

How do we implement MiCA support?

Markets in Crypto-Assets Regulation (EU 2023/1114) requires CASP (Crypto-Asset Service Provider) licensing in one EU state with passporting. Technical requirements affecting development:

White paper is mandatory for issuers of ART (Asset-Referenced Tokens) and EMT (E-Money Tokens)—not a marketing document but a legally binding prospectus with technical description, holder rights, and redemption mechanisms.

Custody requirements: client assets separate from operational assets. Technically: separate wallets/accounts per client (or omnibus with off-chain mapping + regular reconciliation), no possibility to use client funds for operational needs.

Transaction monitoring and reporting: CASPs must keep records of all transactions for at least 5 years and provide them to the regulator upon request.

Travel Rule in MiCA: the threshold for VASP-to-VASP transfers is zero (not the FATF threshold). Implementation requires a Travel Rule endpoint operating 24/7.

Compliance infrastructure implementation process

Compliance architecture is not added on top of an existing product without pain. The correct order: compliance requirements → data model → business logic → UI. If you already have a product without a compliance layer, we start with a gap analysis: what data is already collected, where the gaps are, what will require schema migration.

1. Gap analysis — audit of current architecture and data flow (1–2 weeks).
2. Design — selection of KYC provider, Travel Rule protocol, AML tool, data model.
3. Integration — connecting KYC API, implementing AML screening in the pipeline, setting up Travel Rule gateway.
4. Testing — end-to-end tests, simulating Travel Rule handshake, verifying sanctions screening.
5. Deployment and monitoring — rollout with feature flags, setting up alerting for compliance service errors, audit trail.
6. License support — preparing documentation for the regulator, assisting with inspections.

What does the blockchain compliance service include?

• Compliance architecture documentation (data flow, ER diagrams, API specifications).
• Integration of KYC/AML/Travel Rule APIs with your backend.
• Setup of monitoring and alerting for compliance services.
• Training your team on tools (Chainalysis, Sumsub, etc.).
• Support during the licensing process (MiCA, FATF).

Timeline benchmarks

• KYC/AML integration with Sumsub or Jumio — from 3 to 6 weeks.
• Travel Rule (TRISA or Sygna) — from 6 to 10 weeks.
• Full compliance infrastructure for CASP licensing — from 4 to 8 months.
• On-chain compliance via VC/SBT with ZK (MiCA-ready) — from 5 to 9 months.

Scope is refined after gap analysis. To evaluate your project, contact us—we will conduct a free analysis of your current architecture and select the optimal set of tools. Get a consultation on compliance architecture for MiCA or Travel Rule. Our team has over 7 years of blockchain development experience and 15+ deployed compliance solutions. Request an audit of your protocol for compliance with current regulatory requirements.