Facctum

Alert scoring and the auto adjudication engine use weighted scoring and probabilistic models to resolve screening alerts and learn from past decisions, AI auto tuning adjusts screening profiles, and models turn raw sources such as press releases into screenable profiles. Matching against watchlists would still run without them, so the models carry the costly part of the job, clearing and tuning alerts, rather than the whole product.

Which alerts the engine may close on its own, at what score, and when a case goes to an analyst are not described. The auto adjudication engine resolves alerts, reapplies earlier decisions to recurring alerts and prioritizes high risk alerts under rules the institution sets, with role based access control and an end to end audit trail. Transaction monitoring adds adaptive thresholds, and payment screening lets an institution customize rules and automate decisions.

Accuracy, validation and drift figures for the adjudication and matching models are not published. Each alert decision carries stated reasoning built from weighted scoring, probabilistic modeling and enrichment filters, customer screening returns explainable match decisions from attribute weighted scoring, and detailed logs support regulatory reporting. Transaction monitoring rules are refined continuously, with link analysis and dashboards for investigators.

Customers go unnamed. The figures are Facctum's own and attach to no institution: false positives and manual review cut by up to 70 percent, know your business costs down by up to 40 percent per case, more than 5 million watchlist records managed by clients, potential matches returned in under 80 milliseconds on average and 99.9 percent uptime. Celent has recognized the company among technology providers transforming compliance.

Whether the adjudication models' learning stays within one institution or pools across clients is not stated, nor are training or retention terms for alert data. The models learn from previous outcomes and from each organization's adjudication history, and AI auto tuning adjusts screening profiles from results.

The privacy policy of Facctum Solutions Limited covers its customers and website visitors, while retention, subprocessors and processing terms for the customer and payment data an institution screens are not set out. Screening applies privacy first rules to sensitive data, and an institution chooses the cloud region its data sits in.

An auditor, certificate scope and a trust center are not given. Facctum states ISO/IEC 27001:2022 certification across its operations, infrastructure and products, adds ISO 27017, ISO 27018 and SOC 2, and states compliance with the General Data Protection Regulation and the EU Digital Operational Resilience Act.

Data at rest is encrypted with 256 bit keys in hardware security modules validated to Federal Information Processing Standard (FIPS) 140-2 Level 3, and data in transit with Transport Layer Security 1.2.

Facctum states no regulatory position or admission of its own, as a technology supplier to regulated institutions. Its products are framed around anti money laundering, sanctions and know your customer duties, with payment screening aligned to Single Euro Payments Area rules, the EU anti money laundering directives and Financial Action Task Force guidance, and compliance stated with the EU Digital Operational Resilience Act.

Fairness testing of name matching or of the adjudication models is not published, and no governance process for the learning models is described. Screening decides which customers and payments are held for review, where uneven false positive rates across names, scripts and languages would fall, and AI auto tuning changes match behavior over time. Customer screening includes impact analysis for configuration changes.

A guarantee or indemnity behind an adjudication is absent. What exists is traceability: every alert decision carries its reasoning and an end to end audit trail, so an institution can review and challenge how an alert was closed.

Model providers, the cloud host, watchlist data suppliers and subprocessors go unnamed. Facctum describes the method instead: weighted scoring, probabilistic modeling and enrichment filters inside its own adjudication engine, and models that structure unstructured sources into screening profiles.

Specific core banking or case management products are not named, and full interface documentation comes after contact. FacctView and FacctShield interfaces cover real time and batch screening across more than 20 payment message formats, with webhooks and case management integration, and authentication runs through OAuth 2.0 with single sign on to an institution's identity provider. Watchlist management feeds the downstream screening engines through configurable output templates and automated files.

The cloud provider and any on premises option are not named. The platform is cloud native and available in more than 30 regions, so an institution can choose the one that meets its data residency rules, with multiple deployment options for resilience and disaster recovery behind its 99.9 percent uptime.

Price, tiers and the unit of charge are unpublished, and full interface documentation and a sandbox come after contact. Cost becomes known only in a sales conversation.

Banks, payment service providers, fintechs, neobanks, wealth and asset managers, capital markets firms, crypto exchanges, currency exchange businesses, ecommerce and gaming operators each have a page written for their obligations, and the frequently asked questions add insurers and digital platforms. A buyer in any of those segments finds material on its own screening duties.