What it does
Feedzai markets a fraud and financial-crime platform spanning transaction monitoring, scams, account opening, payment fraud and case operations. The technical value proposition is real-time scoring from transaction, customer, device, behavioral and network signals, combining machine-learning models with rules and investigator workflows. It is not a single model and implementation outcomes depend on available labels, integration latency and the operating policy around scores.
Wio Bank gives the product names an operating context
Feedzai’s undated Wio Bank case, reviewed September 29, 2026, names Digital Trust and Transaction Fraud for Banking as the solutions used by the Abu Dhabi-based bank. It includes commentary from Wio’s Head of Fraud Risk, Jo Jeyaseelan. The inspected page establishes a vendor-hosted customer account but supplies no comparable fraud-loss or series. It should not be turned into a quantified return-on-investment claim. [5]
Analysis: combining information about a digital session with a payment event may help a bank decide when to intervene. The operating result depends on what follows the score: a hold, additional authentication, an investigation or a customer conversation. For an authorized scam, a customer’s confirmation may still be consistent with deception; the intervention needs to match the threat.
Jack Henry is a distribution route, not another bank customer
Jack Henry’s own product-context page identifies its collaboration with Feedzai behind Financial Crimes Defender. Separately, Jack Henry’s February 23, 2026 release says SELCO Community Credit Union is using Defender to bring fraud and BSA/AML work into a shared environment. Together these sources document a route through a banking-technology provider to an institution. They do not establish a direct SELCO–Feedzai contract or that SELCO uses every standalone Feedzai module. [6][7]
In that release, Stephanie Ziegler, SELCO’s Director of Financial Investigations, says Defender provides “a clearer view of our fraud and compliance universe.” This is attributed customer commentary about operating visibility; the release does not provide an independently controlled measurement of losses prevented. [7]
Buying through a platform changes the diligence map
Analysis: a bank evaluating an embedded service needs to know which party supports the interface, supplies model updates, investigates incidents and maintains the audit record. A technology-provider relationship can simplify integration while creating an additional dependency. The bank should understand the route by which an error in an upstream signal becomes visible and correctable in its case system.
Scroll horizontally to see all columns.
| Question | Evidence to obtain | Interpretation |
|---|---|---|
| What was deployed? | Named modules, covered payment rails, live date and traffic share | A platform selection is not proof of every feature being active. |
| What improved? | Comparable fraud dollars, , investigation hours and mature outcomes | Visibility and efficiency gains do not automatically equal loss prevention. |
| Who can change a decision? | Version records, approval authority, rollback and exception logs | A model update and a bank policy change have different owners. |
| What survives an outage? | Fallback behavior, delayed-event replay and reconciliation results | A fast normal path does not establish resilience. |
Actual ML versus marketing
Machine learning can rank risk, detect nonlinear interactions and adapt models more efficiently than static rules. It cannot independently determine legal liability, customer intent or the proper friction threshold. Feedzai public case studies report outcomes such as a 37% improvement in detection for one customer and alert-backlog reduction for another; these are vendor-selected claims. They do not disclose enough common baseline, loss maturation, cost or independent replication to generalize.
Evidence gates
Scroll horizontally to see all columns.
| Gate | Minimum evidence | Why it matters |
|---|---|---|
| Baseline | Current model, fraud rate, approval and review policy | A lift percentage needs a denominator |
| Labels | Fraud, scam, first-party, dispute and recovery definitions | Different labels imply different controls |
| Performance | Recall, precision, , dollars saved and friction | AUC alone does not establish operating value |
| Maturity | Observation window and aged losses by cohort | Early results can miss delayed |
| Governance | Reason codes, versions, overrides, fairness and monitoring | Challenge, audit and controlled change |
| Resilience | Latency, capacity, fallbacks, replay and incident response | Real-time rails leave little recovery time |
| Economics | License, integration, review cost and prevented loss | Adoption does not establish positive ROI |
Implementation burden
A mature deployment requires event schemas, historical labels, streaming integration, identity linkage, case-management design, sanctions/BSA boundaries, and feedback loops. Human-in-the-loop review should have explicit authority, service levels and escalation. Champion/challenger testing should segment by channel, merchant, geography, payment rail and customer tenure; aggregate gains can hide a deteriorating niche.
Sources
- Feedzai RiskOpsSource
- Vendor case study: detection improvementSource
- Vendor case study: modernizationSource
- Feedzai ML overviewSource
- Feedzai: Wio Bank customer story; undated page reviewed September 29, 2026SourceBack to text: ↑
- Jack Henry: Feedzai collaboration behind Financial Crimes Defender; undated page reviewed September 29, 2026SourceBack to text: ↑
- Jack Henry: SELCO Community Credit Union deployment; February 23, 2026; vendor/customer announcementSourceBack to text: ↑1↑2↑3