Shift4 adopts ThetaRay AI to scale AML transaction monitoring

How do static rules block product velocity?

For fast-scaling fintechs and merchant acquirers, compliance cannot act as a blocker to business expansion. When Shift4 prepared to launch new payment products, such as Local Payment Methods (LPMs), its legacy vendor stalled.

• Inability to ingest new data: The legacy vendor admitted its infrastructure could not ingest data formats required for new payment channels, halting product rollouts.
• Slow calibration cycles: Reviewing rules once every few months left the risk team reacting to threats late, incurring heavy documentation burdens.

As Shift4 expanded globally across sports stadiums, e-commerce and merchant acquiring, its existing AML monitoring framework reached a breaking point. Static rule-based systems introduced severe operational liabilities:

• Extreme alert fatigue: An 89% false positive rate meant analysts spent thousands of hours manually querying BI databases to clear low-quality alerts.
• Strict SLA deadlines: Regulatory expectations in jurisdictions like Malta mandate same-day Suspicious Activity Report (SAR) filings, a near-impossible feat when analysts receive alerts late in the afternoon while wading through backlogs.
• The "black box" audit fear: The traditional compliance mindset was built around static formulas and thresholds. Switching to machine learning raised fears about auditor scrutiny and regulatory pushback.

Measure	Before: rule-based	After: ThetaRay AI
Alerts generated	100	20
False positives	89%	12%
Risk data	Manual data extraction	30+ weighted risk indicators
SAR deadlines	High risk of missed deadlines	Same-day SAR filing

Empowering the compliance team with multidimensional AI and explainable risk scoring

In late 2023, Shift4 selected ThetaRay's Transaction Monitoring AI Platform to replace static threshold rules. Full European deployment was achieved in Q1 2024.

When Shift4 developed a brand-new payment product in 2024, ThetaRay completed the data ingestion and custom monitoring architecture in under one quarter, well ahead of product go-live.

Compliance experts from ThetaRay meet weekly with Shift4's risk team to proactively adjust features, threshold materiality and peer groupings.

How does explainable, weighted risk scoring work?

Rather than binary thresholds, ThetaRay analyzes over 30 weighted risk features per transaction, comparing behavior against industry peers, historical client behavior and statistical anomalies.

Every alert displays the key risk contributors and their weighted percentages, enabling analysts to immediately understand why an alert fired, so they are never left shrugging their shoulders during audits.

Noise-cancellation mechanisms suppress repetitive duplicate alerts on the same merchant, firing only when there is a genuine shift in risk behavior.

Results: audit readiness and proactive scale

Within seven months of deployment, Shift4 transformed its risk management posture:

• High-yield investigations: Boosted productive alerts by 70%, shifting compliance culture from clearing backlogs to high-value financial crime investigations.
• Multi-jurisdictional assurance: Met aggressive reporting deadlines (e.g., same-day SAR filings in Malta) with complete confidence during regulatory audits.
• Frictionless growth: Absorbed rapid volume growth across 200,000+ businesses with zero headcount expansion.

Business ROI and scalability

By integrating ThetaRay's AI platform, Shift4 streamlined daily operations:

• Rapid adaptation: When launching new payment channels, ThetaRay's platform adapted quickly to ingest data and monitor new risk profiles without months of rule engineering.
• Instant risk highlights: ThetaRay's AI highlights key risk factors instantly, turning hours of manual research into minutes.
• Weekly feedback loop for model refinement: Weekly syncs with ThetaRay replaced slow, quarterly rule review cycles, allowing continuous optimization.
• 70% increase in productive alerts: Shifted operational focus directly toward actionable investigations leading to SAR filings.
• Superior unit economics: Protected global growth across 200,000+ merchants without expanding compliance team headcount.

Leadership lessons for operations and product leaders

• Treat compliance as a business growth enabler, not a blocker: Deploying adaptable AI monitoring ensures product launches and new corridor expansions are never delayed by data ingestion barriers.
• Decouple revenue growth from headcount growth: Scaling transaction volume by orders of magnitude does not require a linear increase in compliance FTEs when noise is eliminated.
• Embrace incremental operational iteration: Weekly operational syncs with solution partners allow systems to adjust dynamically to changing market risks rather than relying on stale quarterly updates.
• Contextual data visualizations accelerate workflows: Presenting analysts with visual entity paths and weighted features turns hours of manual data extraction into rapid decision-making.
• Avoid the pitfall of custom legacy builds: In-house static rules built over BI tools quickly lack the auditability, case management and agility required for international expansion.
• Focus on total cost of ownership (TCO): Reducing alert volume by 80% and false positives by 86% delivers immediate operational savings that far outweigh modern SaaS implementation costs.

Watch the Shift4 x ThetaRay webinar on Cognitive AI in transaction monitoring.