Business Expertise
Project Implementation
Project Implementation Consulting
Support Services
SWIFT Related Applications
Transaction Pattern Analysis


CubeIQ Solutions and Systems

Welcome to CubeIQ products and systems web pages.

CubeIQ Products and Systems

Group: Enterprise Software Applications


Risk and Compliance Solutions

Anti Money Laundry (AML) Software Applications

Transaction Pattern Analysis


Transaction Pattern Analysis in SmartAML™

There are three major steps in the transaction pattern analysis:

Data gathering (ETL).

Implementation of algorithms.

Results Processing and Investigations.


Data gathering with ETL: Extraction, Transformation, Loading

We are supplying the necessary tools to load in a transaction database all elements found in payment transactions:

SWIFT transfer messages.

SWIFT transactions involving amounts although not transferred.

Domestic transfers.

Cards Transactions from POSs, ATMs, Internet, other alternative channels ...

Basic transactions (deposits, withdrawals, ...).

There are facilities to select a subset of the data by using many different keys, such as period, countries, individuals, currencies, etc…

Data selected are then entered regularly in datasets where each is analyzed according to patterns.

The same data may be entered in several datasets. Each data keeps a link to the original data.

Every individual encountered is created as an account; there will be facilities to group accounts identifications as synonyms, even in different currencies. Amounts are loaded in their equivalent in the reporting currency.

The main data database can also be loaded with ancient or archived data: the datasets can be re-evaluated as and when wished, for instance when an algorithm is modified to reflect an additional element.



We are building terms of algorithms like a collection of statistical or mathematical items:

Average balance of transactions in a given period (day, week, month, …).

Maximum amount transferred.

Number of transactions per period.

Absolute value of transactions per period.

Then you can name complex algorithms to be applied to particular sets of data.

The system is supplied with few algorithms in order the user to understand their building. Also a number of samples are provided such as :

Account that receives and pay above X per day.

Account that enters and exists very quickly everyday, and has a very big volume in the same day, but a small End - Of - Day (EOD) balance.

One account receiving most of amounts from another account.

One account receives foreign funds and disperse to same list most of the time.

Dormant account becomes active with big amounts.

Account becomes dormant after massive funds dispersion.


Result processing and investigations

We are supplying the necessary tools to load in a transaction database all elements found in payment transactions.

Each algorithm can have its alarm parameters, like which report to produce and to whom it should be addressed.

Each instance can have different parameter set.

As an instance run the pattern analysis on a customer from a branch the outcome could be returned to the branch manager as well as the compliance manager.

When a suspected pattern is detected, it becomes a case to be further analyzed: it is then stored in a case database, together with its details assembled in a delimited file directly loadable in Excel ™ for detailed analysis and decision making, including full details of originating transaction.

From there, the officers can analyze it, create follow-up memos or messages, and assemble additional evidences to the case constructed.


For more information please visit:


Swallow Technology (SwallowTech) Ltd.


Related products:



For any question or comment please contact our support team at support@cubeiq.gr 

For product or services inquiries please email to: sales@cubeiq.gr



















News Archive






Site Map


Contact Us




























Term Of Use


Privacy Policy