-
151
Projects submitted
-
109
Published projects
-
47
Comments left
-
66
Number of votes
-
52
days until voting ends
Data factory for Kazakhstan's national payment operator
-
Project manager on customer side
Adilbek Sagadiyev
The National Payment Corporation of the National Bank of the Republic of Kazakhstan (JSC "NPC")
Chief Architect of Technical Solutions
-
Category
-
Nomination
-
Goals
In 2025, The National Payment Corporation of the National Bank of the Republic of Kazakhstan (JSC "NPC") focused on data as an additional source of business value. However, analytics were fragmented: each product had its own system, supporting existing solutions was expensive, and there was no unified data processing platform. A separate requirement was to work with sensitive information: banking secrecy and personal data. NPC was looking for a technology that could easily scale with about 8 TB of data per quarter, support S3 object storage, and combine a data lake and data warehouse in one system.
-
Timeline
march, 2024 — november, 2025 -
Project completion year
2025
-
Project scale
4800 human-hours -
Results
Selena Lakehouse became a key element of NPC's data factory. The corporation saw the following business outcomes for the project:
reduction of infrastructure costs — up to 65%;
reduction of costs for fragmented analytical systems — up to 60%;
reduction of data duplication — up to 20%;
reduction of time for building analytical reports by 4 times.
The implementation of Selena Lakehouse forms the technological basis of NPC's unified data factory. Centralization of analytics and reduction of information duplication create conditions for cost optimization, development of digital services, and effective data management.
-
Project uniqueness
The project became the first Data Lakehouse implementation by a government-owned organization in Kazakhstan. Data Lakehouse approach helped to build an analytical foundation across previously isolated products while handling sensitive financial and personal information. It brings open table and file formats, object storage and an SQL engine together, allowing multiple analytical use cases to share a data foundation. The design combines product analytics for remote identification with an antifraud analytics scenario, rather than building another isolated reporting system. The project is distinctive for its combination of cross-product consolidation, sensitive-data constraints and payment-sector scale. The architecture and results remain subject to customer confirmation.
-
Software used
Selena Lakehouse and Selena Engine; SQL; Apache Iceberg; S3-compatible object storage; Parquet, ORC and CSV; data ingestion/export and ETL tools; BI and reporting.
-
Solution from the Global CIO catalog
The project does not use solutions from the Global CIO catalog
-
Project implementation complexity
The central challenge is consolidating analytics from previously separate products into a common data platform while accommodating banking secrecy and personal data. The use cases span identification and antifraud, with different sources, analytical needs and stakeholders. Integrating ingestion, open formats, object storage, a catalog and SQL-based reporting requires coordinated architecture and operational controls. The presentation does not specify the exact security controls, migration steps, availability commitments or deployment configuration; these require confirmation.
-
Description
Selena receives data from sources via connectors and built-in loading tools, stores it in S3 in open formats — Parquet, ORC, CSV, and Apache Iceberg — and serves SQL queries, ETL, BI, and reporting from a single entry point. NPC notes the built-in data caches, high performance — according to the company, especially compared to Trino-based lakehouse solutions — and an AI assistant for writing queries.
One of the first scenarios was product analytics for the Identity Data Exchange Center (IDEC) — a multi-vendor platform for comparing facial images. The service processes over 2.5 million requests per month, and banks, non-banking organizations, and 8 biometric solution providers are connected to it. IDEC events — from the start of authorization and video identification to electronic signatures — are collected in Selena, where analysts build queries and dashboards based on them.
-
Project geography
NPC ensures the operation of Kazakhstan's national payment system: 89.5% of the country's payment turnover — 5.9 trillion tenge and about 350 thousand transactions per day — passes through the corporation's systems. In addition to interbank settlements, NPC develops digital services: remote digital identification, the Digital Tenge platform, Open Banking, and the Anti-Fraud Center.
-
Customer
The National Payment Corporation of the National Bank of the Republic of Kazakhstan (JSC "NPC")
-
IT provider
Deonix
Only authorized users can comment.
We suggest you log in or register.