Callsign

Identity is inferred from how a person types and swipes, combined with device fingerprinting and ensembled in the Intelligence Engine. Only an ordinary multi factor prompt would remain, the very friction the product exists to remove.

Authentication runs automatically and invisibly, and when social engineering is detected the customer still makes the call after a contextual warning. Thresholds, what follows when a customer proceeds anyway and escalation to someone at the bank are not described.

Detection rises 40 percent while acceptance holds at 99 percent, a form that shows real improvement rather than a shifted threshold, and telecommunications signals lift detection to 70 percent. A 99.999 percent precision claim for behavioral recognition carries no method, sample or definition.

A major United States bank and a tier one retail bank are quoted without names, the second saying customers noticed nothing different. Detection figures are self reported and tied to no named institution.

Exposure improves behavioral models, and the platform serves competing banks while fusing telecommunications intelligence that spans them all. Logical separation of client datasets addresses storage rather than learning, and whether one bank's patterns shape another's detection is not stated.

Behavioral biometrics means continuously observing how a person types, swipes and holds the device, yet no retention schedule, subprocessor list or data processing terms for behavioral profiles are published. The privacy notice is a separate document.

SOC 2 Type II under the American Institute of CPAs standard is named, alongside AES 256 encryption at rest and backups tested twice a year. ISO 27001 is described as alignment on one page and listed as held on another, and no report is downloadable.

PSD2 and Strong Customer Authentication, plus know your customer and anti money laundering requirements, are the named regimes, and the Payment Systems Regulator's reimbursement rules for authorized push payment fraud are discussed. Supplier licensing is neither claimed nor needed.

Digital Trust Index research describes exclusion across the population, not the product's own performance. Behavioral biometrics varies with age, motor control and disability, so atypical typists face more challenges, and no breakdown across those groups appears.

Step up authentication replaces silent denial, and a customer about to send money to a scammer is told what the system has seen. Recourse for someone wrongly challenged or blocked is not described, and no guarantee or indemnity is published.

In house models handle behavioral, device and ensembling work, and carrier signals are disclosed as an external source by type. Carriers and any subprocessor list are unpublished.

An orchestration layer controls which authentication or intervention applies to each journey, and telecommunications operators feed network signals showing a customer is on a live call during a payment. Its documentation portal names no banking platform or identity system as a target.

AWS hosts subscriber data across three geographically dispersed availability zones, with no region or residency commitment named. Profiles of this kind count as sensitive personal data in several markets across the United Kingdom, the United States and Asia.

Usage based licensing by users, transactions or integrations appears only in analyst documentation, not on the vendor's own site, and a demo request is the way in. Without a published plan or basis of charge, cost cannot be sized in advance.

Account login, online payments and account creation are the journeys covered, with account takeover, SIM swap and synthetic identity among named fraud vectors. Banks are the stated buyers, yet none is named.