In order to develop projects that make our customers’ life easier, who are large data users in different sectors where we actively provide consultancy services, we are feeding our R&D projects with our consultancy experience and carrying out studies in the following issues.

• • Anomaly (Outlier) Detection
• • Statistical Learning
• • Telecom / Banking Data Warehouse Model
Our Products
Banking Legal Data Warehouse Data Model

It is the formation of a financial data model to deliver the banking data from different source systems (Accounting, Deposit, Loan etc.) in the form that is consistent with each other in sections related to departments to be created for legal authorities (BRSA, MB, TURKSTAT, Ministry of Finance, Risk Centre etc.), within the requested time interval and with high quality and free from problems.
In addition, it is compulsory to transform the recorded data according to in-bank definitions in the bank's operational system into a form in accordance with definitions of the legal institution. Since reporting medias of institutions are often created based on in-house definitions, the absence of data ready to be reported to official authorities or formation of them with manual operations are very risky in terms of Legal Reporting obligations.

Since the centralized quality data, generated with reference to the Official Authority definitions, are verified in terms of content, no further correction costs will be required. In this respect, 200 reports which are submitted by different business departments will be able to be reported consistently and within the requested time interval, minimizing both the financial risk and the work-force effort spent to make the notifications correctly.
Data Quality Scoring Related To Machine Learning

In sectors such as telecom, banking and insurance where large data are included; it is the product determining data quality problems, scoring data quality, improving data quality processes, and providing automatic quality control solutions in the process of obtaining, transforming and making data meaningful. Unlike traditional data quality solutions that run by checking thresholds defined by users, it is a new-generation data quality solution that learns the behaviour of the data over the data history and reports abnormalities that occur outside the routine behaviour of data.

The product also provides input to the Banking Data-Warehouse Conceptual Model, which is our other R&D product, along with the quality scores it produces. In this respect, financial and reputation losses experienced in compulsory notifications made to legal institutions, arising from data quality are avoided.