A data processing platform in TBC Uzbekistan (payme)

A project was set to develop a data platform. The solution integrates internal and external data sources and operations, and enables each department to easily access accurate, time-relevant data. This allows the organization to significantly reduce the time required to make business decisions on various issues.

As a result, a data platform was formed for quick analytical responses, including in self-service mode. The new approach made it possible to reduce time-to-market from several days to one day, in some cases from 24 to 3 hours. Modern technologies ensure the identification of new analytical metrics for products and customers, the formation of individual proposals, tracking the “life cycle” and patterns of customer behavior. The current date of the platform will allow the introduction of several more advanced analytics tools, including data models, in the near future.

Data integration - Debezium, Kafka.

Data core - ETL, Airflow.

Data platform - Greenplum database.

BI report - Power BI.

The main difficulties of the project were a large number of different data sources, and a low degree of documentation.

As part of the project, multiple business data marts were established to enhance data management and reporting capabilities. New data sources were integrated, and custom reports were developed to enable rapid tracking of changes. The data marts provide detailed insights, including key customer attributes and behavior patterns within the payment application, offering a comprehensive view for improved decision-making.