Onboarding – Quickly tracing any Entity by “Name-Sound”, based on a Single Mathematical Representation.
Ongoing – Potential Flags, (similar phonetic Entities or other Attributes) are saved in the systems’ database for Ongoing Verification, thus Eliminating the Risks of Whitelisting of Alerted Entities.
Transactions – Realtime Transaction Screening. Real-time check of ALL Entities involved in a transaction: Payer-Beneficiary-the 2 Banks.
COP/BAV – adding the powerful ability of ‘Accurate Name Matching’ across 38 languages, for confirmation that Beneficiary == Account owner.
a Plugin add-on to existing screening solutions;
Multilingual Screening of unstructured Names or Entities.
Ensuring no False Negatives (Misses).
Fincom.co proven technology, enhances the performance of all current AML Name Screening solutions, adding the unique ability in understanding over 38 different languages, and complex name structures.
This service ensures highest level of compliance even with legacy solutions.
Advanced ML/AI Engines - easily integrated into the existing AML systems as well as a natural part within the work-flow process of the Compliance Team
Reduce the dependency on manual processes. Ensuring consistent outcomes, increasing efficiency, significant reduction in operational costs and risk of errors. Also – eliminating the inevitable inherent Discrimination by name alone, as is in legacy systems (example: Mohamad)
Encompass L2 team’s holistic investigation of the customer, transaction activity as well as other involved entities. Providing Entity Type match, Entity Enrichment screening by Attributes (ex. Address, BD, Geo) , ML Trend Analysis + Historical data, and AI Risk Scoring engine.
Micro Service – Big Optimization
Micro Service
– Big Optimization
Banks are growing whilst being forced to meet the growing compliance regulation requirements; as a result, increase in compliance team size, workload and operational costs.
Alerts are resolved by the compliance team, taking time and effort. This huge investment is not kept for future referencing and verification.
Even worst, past Alert Names are often forwarded to a Whitelist, exposing to major risks of missing real Alerts.
The complex task of screening the customer roster, results in many False Positives that re-appear and requires manual alert-suppression, causing back log and heavy workload requirements.
Many FIs have adopted the methodology of “Whitelisting”, whilst ignoring the fact that whitelisted customers can in time become sanctioned, as more information is surfacing (e.g., ING Bank – 700 Million €, EU fine for whitelisting, 2019)
PHONETICS
– understanding the “sound”
COMPUTATIONAL LINGUISTICS
– convert Sound into Number
DISTANCE ALGORITHMS
– Increase Matching Accuracy
4 Propriety distance algorithms, 2 intra-language & 2 inter-language
UNDERSANTING NAMES
– Enhancing human Eye & logic
REGRESSION ANALYSIS
– persistence (using history of past decisions)
FREQUENCY WEIGHING
– frequency of similarities between names
FUZZY LOGIC
– ensuring efficiency and effectiveness
SUGGESTION ENGINE
– Trend Analysis (combined with
ML-Persistence mechanism)
Using advanced phonetics, artificial intelligence (AI) & fuzzy logic in a multi-dimensional process:
SANITIZATION: ensuring the integrity of the information in the databases
NORMALIZATION: formatting & transforming data to a mathematical code (phonetic fingerprint)
CALCULATION: matching the entries between databases by over 44 algorithms accurately and efficiently.
Phonetic fingerprint is a mathematical representation of the pronunciation and phonemes within the name (individual person or other entity). Using automated Real-time technology, supported by computational linguistics, AI-ML layers, using over 48 mathematical algorithms, tracing the phoneme across 38 different languages, transliterations and spelling variations enables accurate name matching.
An automated process, designed for ongoing verification resolving 60% of Potential Alerts, Automatically!
On each time there is a change in Sanction and PEP lists – changes will be verified automatically, and only real-alerts will be forwarded for further resolution; so unless there is a change of data the Compliance team will not be Alerted for no reason, whilst if there is a change it will not be missed.
provide a wider name-screening coverage. Using advanced phonetic based ‘name-understanding’, enabling seeing through different ways of writing names.
Indexing sanction lists/PEP/Blacklists based on advanced Phonetics enabling rapid and accurate matching. using unique mathematical representations to ensure high-accuracy for both FP reduction and avoid FN.
The use of Phonetic based methodologies (rather than rule based), enables Comparison of Entities in a very wide funnel; more name combinations, languages, transliterations.
The ability to trace ad identify the same name across 38 languages and different pronunciations
I) PHONETIC THRESHOLD – checks name composition; i.e., how many naming words are included in the entity-search; II) NAME-DISTANCE THRESHOLD, resolving the difference between names that spelled or sounds alike.
applying AI/ML technologies. Looking at the frequency of Names, weighing and differentiating common names, unique names or combinations of names.
AML Regulatory compliance demands: Explainable and Traceable screening solutions.
The Optimum system for that is based on pure mathematics – Objective, Traceable process, and completely Explainable. (in difference to rule-based empirical methodology -“what if” which is Subjective).
Fincom.co uses Mathematics in Linguistics Computation and Phonetics formulae, in a way that is explainable and traceable. hence 100% Transparent supporting regulatory compliance auditing, management and adherence.