Lucinity

Luci, the agent, gathers evidence, runs adverse media checks and drafts narratives and regulatory reports, with AI detection now beside scenario based monitoring. A case management tool, minus the capability it is bought for, would be left.

Escalations, thresholds and suspicious activity reports all need the institution's approval while Luci gathers evidence and drafts, so governance stays where it was. Under the managed service, Lucinity's own analysts complete each case Luci prepares before returning it, and every step carries its reasoning, evidence and audit trail. Those named approval points place staff at every consequential step.

Retrieval with a validation step ties statements to case documents rather than model recall, a named architecture a model risk function can assess. Accuracy, false positive and summarization fidelity figures are absent.

Visa Cross Border Solutions, formerly Currencycloud, saves five hours of analyst time a day, and a Microsoft case study reports investigation time cut from three hours to 30 minutes. Finshark is also named as a customer.

Enterprise terms on the model service behind Luci exclude customer prompts and documents from provider model improvement, and case material is retrieved at query time rather than absorbed into model weights. Retention terms behind that arrangement, and any learning across the customer base, are not addressed.

Transaction data, investigations and suspicious activity reports in the platform are not described in the privacy notice, which stops at website visitors and their forms, with rights under the General Data Protection Regulation (GDPR) and erasure once data is no longer needed.

SOC 2 Type II and ISO/IEC 27001:2022 are named in Lucinity's trust center, where the SOC 2 report, the ISO certificate and a 2026 penetration test are released on request rather than for download. The same trust center lists the controls behind them, from restricted access to production systems to set data retention procedures and deletion of customer data when a client leaves.

Suspicious activity report generation and submission make the product regulatory in function, with the money laundering reporting officer as the usual buyer. Supervisory bodies, reporting rules and filing channels are not cited.

Ethical AI is a pledge without a mechanism, while adverse media checks misfire on common names and faster report drafting produces more accusations that follow people. Summaries condense what the deciding investigator sees, and summarization fidelity and population level error testing are unpublished.

Detailed audit logging shows how a conclusion behind a regulatory filing was reached. Whoever a report concerns has no described way to see or contest it, and nothing in the published terms underwrites the output.

Lucinity names its subprocessors and what each one does: Microsoft Azure for infrastructure hosting, Microsoft Azure OpenAI Services for the AI behind Luci, Cloudflare for content delivery and network security, Pendo for in app analytics, Plain for customer support and Salesforce for sales. The adverse media sources behind each check are not tied to a provider.

Partner modules carry their own names: Resistant AI for behavioral monitoring, Sift for payment fraud screening, Facctum or Neterium for watchlist screening and Creditinfo for localized politically exposed person screening.

Luci works as a plugin inside existing workflows, the platform is system agnostic, and its open interfaces come with developer documentation. Partner modules add Sift for real time payment fraud screening, Resistant AI for behavioral monitoring and Facctum or Neterium for watchlist screening.

Oracle has embedded Lucinity's investigation and case management technology in its Financial Crime and Compliance Management platform, so the same capability reaches institutions inside a system they already run.

Azure is the named cloud, with no region, data location commitment or private deployment option. Entities in Iceland, the United Kingdom and the United States serve banks with differing expectations about where crime data rests.

Lucinity publishes no price or basis of charge, whether per investigation, per seat or per alert, and its managed service runs under service level agreements that carry no published rate. An Azure Marketplace route lets a bank pay from existing Microsoft cloud commitments, which can shorten purchasing.

Banks, payment institutions and fintechs are served across transaction monitoring, investigation, adverse media and regulatory report submission. The domain is financial crime operations rather than the wider institution.