11
Shufti, the Glocal identity verification and full compliance lifecycle provider, announced the launch of Transaction Trust Monitoring (TTM), an agentic FRAML engine that carries a case from the first suspicious transaction through investigation to a regulator-ready report, in one workflow. TTM scores each transaction against the identity Shufti verified at onboarding. It then routes alerts through a role-gated investigation process and generates a report in the format required by the Financial Intelligence Unit for the Money Laundering Reporting Officer to review and approve. It integrates with existing KYC and case management systems rather than replacing them. FRAML brings fraud and anti-money laundering controls into a single discipline. In TTM, both sets of signals are scored in the same pass, so a pattern spanning both typologies is assessed once against one customer record.
Where Transaction Trust Monitoring Differs
· Unifies KYC and KYT on one in-house stack: Identity verification at onboarding and ongoing transaction monitoring run on a single platform Shufti owns end-to-end, with no acquired modules and no third-party stitching, so each customer carries one risk score and one audit trail.
· Identity-to-transaction continuity: Each transaction is scored against a biometrically verified customer, with deepfake and synthetic identity defences at onboarding, and behavioural biometrics and device fingerprinting after login, all inform that score. · Screening depth behind every decision: More than 4,000 watchlists and 6M+ PEPs with sanctions coverage across OFAC, OFSI, EU CFSP, DFAT and the UN Consolidated List, and multilingual contextual adverse media.
· 1,600+ data points per transaction, scored in under 500ms: Every transaction is checked against location, device, banking, and threshold data, plus more than 15 velocity checks running from one minute to 180 days, and comes back with one risk score from 0 to 100.
· AI explainability, audit-ready at every step: Every alert carries a case summary, a tier classification, a false-positive assessment, and a recommended next action. · AI rule Builder, live now: Describe a rule in plain language, and Shufti builds it, calibrated by industry, regulation, and risk appetite, then backtests it against 90 days of history to show projected alert volume, precision, and fraud caught by value.
· Alerts routed by score under role-based control: Scores of 75 and above go to the MLRO, 50 to 74 to L2 investigation, and 25 to 49 to L1 triage, each stage carrying its own service level. Flagged transactions can be placed on a soft hold and sent to an analyst queue for review rather than being automatically declined. · MLRO authorisation on every filing: The AI triage co-pilot drafts narratives and recommends actions and does not execute a filing itself.
· Built-in controls on every rule and every alert: Twenty-four ABSOLUTE rules always trigger and cannot be suppressed. Every decision is held on a five-year immutable audit trail, and unscoreable transactions return a NOT ASSESSABLE verdict.
Watch the video below to see how Transaction Trust Monitoring works in practice. https://www.youtube.com/watch?v=G-beTCoCK54
In addition to all, behaviour is also measured against each customer’s own pattern, using a rolling 90-day baseline across 16 dimensions, while linked devices, IP addresses and contact details expose multi-accounting and money mule networks.
Four AI agents are live across the lifecycle, with three more covering rule recommendation, data readiness and regulatory change monitoring in development. MCP integration allows external AI tools to drive a firm’s monitoring setup.
“Compliance teams are covering more jurisdictions and more volume without proportionate headcount, and the work that does not scale sits after the alert,” said Frayam Asif, Chief Technology Officer at Shufti. “We designed TTM where agents score fraud and AML risk inside the payment path, with checks completing before settlement and an analyst opens a case that is already prepared. For our clients, that means less time assembling a case and more time deciding on it, with no added friction for the customer.”
Reporting closes the workflow. Suspicious activity and suspicious transaction reports are generated in the regulator’s native XML, including FinCEN BSA XML 2.0 for United States reporting and goAML, the UNODC schema used by Financial Intelligence Units across Europe, the Middle East, Africa and Asia Pacific. The MLRO reviews, edits, and authorises each report, after which the firm downloads and submits the file with no manual XML assembly or re-keying.
Each filing is logged, with case data, analyst notes, and score explanations available for examination. Threshold reports are obligation-based rather than suspicion-based, so they auto-batch daily under a deadline timer, with the MLRO retaining oversight and able to pause any batch.
The TTM launch is featured in the Shufti Innovation Drop, offering a closer look at its capabilities through a live product demonstration. Watch the full Transaction Trust Monitoring demo here.
Availability
TTM is available now to banks and financial institutions, payment service providers, remittance businesses and MSBs, fintechs and neobanks, virtual asset service providers and iGaming operators, over a RESTful API and SDK, with sandbox access and on-premise deployment. Detection covers SWIFT and SEPA flows, cumulative player history, corridor and agent-level monitoring, and wallet screening, on-chain risk scoring, and Travel Rule support on the same engine as fiat, across 240 countries and territories. Transaction Monitoring Software can be configured as a standalone solution or as part of the complete Shufti compliance lifecycle, from onboarding through ongoing monitoring. Shufti has also published A Guide to Risk-Aligned Transaction Monitoring, which sets out the FATF requirements for monitoring, the five principles the HKMA recommended for selecting a system, and the framework changing under Regulation (EU) 2024/1624 from July 2027.
Read the guide to understand what risk-aligned transaction monitoring looks like in practice. Or contact the Shufti team to request a demo.