FinInsight AI investigates every transaction Safer Payments flags and writes the justification, blocks bot logins and catches edited documents, while iVerifier eKYC proves who your customer is.
Three capabilities share one risk view across digital channels, identities, documents and transactions, with iVerifier eKYC confirming the person behind each session. FinInsight AI and iVerifier work with Safer Payments and sit on top of your existing AML, transaction monitoring, KYC and CRM systems; they don’t replace them.
Telemetry from web and mobile banking stops bot logins, emulators and compromised devices; iVerifier eKYC confirms the real person with face match and passive liveness.
Prioritise real risk, trace mule networks across accounts, and draft audit-ready SARs and STRs with AI.
Check payslips, bank statements and identity documents for signs of editing before an account or loan is approved.
to a ready-to-file SAR/STR draft, from ~2 hours
estimated annual savings for a mid-size bank
Figures reported by FinInsight AI. Results depend on your data, alert volumes and processes.
Safer Payments monitors transactions in real time and flags suspicious ones, but it doesn’t dig deeper into each flag, so false positives land on your analysts. FinInsight AI connects through APIs on both sides: it adds device context going in and investigates every flagged transaction coming out.
Device Intelligence passes session telemetry and behaviour signals to Safer Payments, so each transaction is scored with what happened on the device first.
FinInsight AI pulls flagged transactions, gathers evidence from core banking, KYC, CRM and AML systems, and its AI model separates false positives from genuine risk.
Generative AI writes a source-cited justification for every cleared alert and drafts the case narrative and STR when risk is real. Analysts review and sign off.
No rip-and-replace: Safer Payments keeps making real-time decisions. FinInsight AI takes over the investigation and write-up work that analysts would otherwise do by hand.
Device, browser, network and session signals from web and mobile banking are scored for every session, so bots, emulators and compromised devices are stopped at login.
Borderline sessions and high-risk changes, such as new device binding or contact-detail updates, get a live selfie matched to the ID with passive liveness.
Run document verification, biometric matching and liveness, the three checks in BNM’s e-KYC policy document, in one hosted session with scores, reason codes and a signed verdict.
FinInsight AI is the AI-native layer for AML investigation teams: it picks up alerts from Safer Payments and AML screening, prioritises real risk, traces hidden networks across accounts and drafts audit-ready SARs and STRs, so investigators spend their time on the cases that matter.
Flagged transactions, KYC, alerts, documents and case notes pulled from your banking systems into entity-level timelines and graphs, with long-term behavioural memory across cases.
Risk heuristics such as transaction profile, customer risk tier and alert repetition resolve low-risk alerts automatically, cutting false positives by up to 50%.
Explainable case summaries, investigative rationale with source-level reasoning, and ready-to-file STRs with built-in validation.
Sanctions, PEP and adverse media screening, sub-second transaction risk evaluation, and thresholds tuned to each customer risk tier.
Deep linkage across customer, account and transaction entities surfaces hidden relationships and mule clusters that static rules miss.
Every flag ships with a human-readable, source-cited rationale, not a black-box score, and the engine keeps learning from investigator feedback.
On-premise, SaaS or hybrid; API-first into Safer Payments and your AML and transaction monitoring stack; zero PII persistence, full audit trails and role-based access. Most banks go live within a few weeks.
Edited payslips, altered bank statements and doctored IDs are used to open accounts and obtain loans. FinInsight AI checks submitted documents for signs of tampering before approval, and sends anything suspicious to an analyst or an AML case.
Screen payslips, bank statements, identity and business documents as they arrive in onboarding and loan cases.
Each flag explains what looks wrong, such as edited figures or inconsistent details, so analysts can decide quickly.
Results are filed with the case in Alfresco, and suspected fraud can open an AML investigation with the evidence attached.
Pair it with iVerifier eKYC: when the document is an ID, iVerifier also checks that it is genuine and that its holder is present, with face match and passive liveness.
FinInsight AI works best on a platform that keeps evidence, cases and APIs under control. Crest delivers each part on its own, and they fit together.
Alfresco Enterprise: federated documents and records with retention, audit and AI search, self-hosted for residency.
Onboarding, loan and card cases with SLAs, exceptions and evidence on Alfresco Process Services.
SignDex: PAdES-LTV signing inside your workflows, with every signed artefact filed as a governed record.
Kong Enterprise: mTLS, OIDC, OAuth, rate limits and observability for every channel and partner API.
iRetailOS: appointments, virtual queues and waiting-time visibility for every branch.
Terraform Enterprise, Vault Enterprise and StackGuard: codified infrastructure and short-lived, rotated credentials.
Crest Infosolutions designs, implements and supports these platforms for regulated organisations from offices in Malaysia, Singapore, Indonesia, India, the UAE and the USA.
Plan fraud, AML, identity and data controls together, against the frameworks your examiners and auditors use.
Compliance depends on your policies, configuration, operating procedures and audits. These platforms provide controls and evidence that support your compliance programme.
AML/CFT rules, such as Malaysia’s AMLA 2001 and MAS Notice 626 in Singapore, require banks to screen, investigate and report suspicious transactions. Anti-scam expectations are rising: Singapore’s Shared Responsibility Framework sets duties including real-time fraud surveillance, and Malaysian banks have added single-device registration, cooling-off periods and kill switches. BNM’s e-KYC policy document and RMiT policy, the MAS Technology Risk Management Guidelines, PCI DSS v4.0.1 and the PDPA, whose 2024 amendments add biometric data to sensitive personal data, shape identity, technology and data controls.
FinInsight AI: source-cited rationale, audit trails, role-based access and zero PII persistence for AML decisions and STR drafts. Safer Payments + FinInsight AI: a real-time decision trail plus a written investigation trail from flag to filed STR. Device Intelligence: session-level evidence for anti-scam controls. Alfresco: retention, classification and audit for case evidence. iVerifier: document verification, biometric matching and liveness for e-KYC. Kong: API authentication, rate limits and logs. Vault and StackGuard: short-lived, rotated credentials with no hard-coded secrets.
Practical answers for fraud, financial crime, compliance and technology teams.
FinInsight AI is Crest’s AI layer for fighting fraud and financial crime in banks. It has three modules: Device Intelligence for web and mobile banking, AI-driven AML case management with SAR and STR drafting, and document fraud detection. It works with Safer Payments and sits on top of your existing AML, transaction monitoring, KYC and CRM systems.
It collects telemetry from your internet banking site and mobile app, detects suspicious browsers, emulators and compromised devices, and scores every session in real time. High-risk visitors can be blocked, or stepped up for extra verification, before they reach payments or account changes.
Safer Payments monitors transactions in real time and flags suspicious ones. FinInsight AI adds device and behaviour context for those decisions, then pulls each flagged transaction through APIs, gathers evidence from core banking, KYC, CRM and AML systems, and uses an AI model to separate false positives from genuine risk. Generative AI writes the justification for every outcome and drafts STRs for cases that need escalating.
No. Safer Payments or your transaction monitoring solution keeps making real-time decisions. FinInsight AI adds the investigation and case-management layer on top, connected through APIs, so there is no rip-and-replace. Most banks go live within a few weeks.
FinInsight AI brings transactions, KYC, alerts, documents and case notes into one investigation context, then applies risk heuristics such as transaction profile, customer risk tier and alert repetition to resolve low-risk alerts automatically. It reports up to 50% fewer false positives.
Yes. It generates explainable case summaries and investigative rationale with source-level reasoning, and drafts ready-to-file STRs with built-in validation. Investigators review and file them; FinInsight AI reports drafting time falling from about two hours to about 12 minutes.
FinInsight AI checks submitted documents for signs of editing, anomalies in metadata and structure, and inconsistencies across documents and the application, before approval. Each flag comes with reasons, so analysts can clear, review or decline, and suspected fraud can open an AML case.
FinInsight AI spots risky sessions, documents and accounts; iVerifier confirms the person. Use it at account opening, and as a step-up check when Device Intelligence or an AML case raises risk: one hosted session checks the ID document, matches a live selfie to it and runs passive liveness, then returns a signed verdict to your case.
FinInsight AI runs on-premise, as SaaS or hybrid, with zero PII persistence, full audit trails and role-based access. Crest lists SOC 2 Type II, ISO 27001, VAPT testing and GDPR compliance for FinInsight AI.
Crest also delivers the platform underneath: Alfresco document and records management and case management, iVerifier eKYC, SignDex signing, Kong API gateway, iRetailOS branch queue management, and Terraform, Vault and StackGuard for infrastructure and secrets.
Bring anonymised alerts, a document sample or a channel to protect, and we’ll show you the result in a 30-minute walkthrough.
Telemetry · detection · bot blocking
iVerifier · document, face match, liveness
Investigate · justify · draft STRs
Prioritise · trace networks
Payslips · statements · IDs
Alfresco · Kong · SignDex · Vault
Request for information or schedule a solution demo.