Lakehouse Architecture for Financial Systems: A Unified Approach to Compliance, Analytics, and Risk Management

Financial institutions have long operated between two inadequate extremes: the data warehouse, which offers governance and reliability but resists the scale and flexibility that modern analytics demand, and the data lake, which provides flexibility at the cost of structure, trust, and regulatory traceability. The Lakehouse paradigm resolves this tension by layering transactional semantics, metadata management, and open storage formats directly onto cloud object storage. This paper proposes a six-layer Lakehouse reference architecture for financial systems, grounded in the Banking Industry Architecture Network (BIAN) Service Landscape 12.0 service-domain model. Each layer is mapped to the service domains that own its data products, which gives the platform a clear line of accountability. The architecture addresses ingestion and source integration, open table storage using the medallion pattern, governance and compliance, analytical and risk serving, real-time streaming integration, and data product delivery. A structured comparison of Delta Lake, Apache Iceberg, and Apache Hudi across financial workload profiles (batch regulatory reporting, real-time risk scoring, and ML model training) guides open table format selection decisions. Design tensions between real-time processing latency and regulatory auditability are analysed, together with a federated governance model that reconciles domain ownership of data with the enterprise-wide lineage that BCBS 239 requires, and a phased implementation roadmap is provided. The framework provides data architects and platform engineers with a structured foundation for building unified financial data platforms that satisfy BCBS 239, GDPR, SOX, and EU AI Act obligations while delivering the analytical throughput and ML serving capabilities that modern risk management demands.

Authors

Publication Details

Journal
East African Journal of Information Technology
Published
2026-11-14
DOI
https://doi.org/10.37284/eajit.9.3.6281
Primary Topic
Big Data and Business Intelligence
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Lakehouse Architecture for Financial Systems: A Unified Approach to Compliance, Analytics, and Risk Management

Ratanchur Sarkar, Sovon Nath
East African Journal of Information Technology
Big Data and Business Intelligence
article

Lakehouse Architecture for Financial Systems: A Unified Approach to Compliance, Analytics, and Risk Management

Ratanchur Sarkar, Sovon Nath
article en

Abstract

Financial institutions have long operated between two inadequate extremes: the data warehouse, which offers governance and reliability but resists the scale and flexibility that modern analytics demand, and the data lake, which provides flexibility at the cost of structure, trust, and regulatory traceability. The Lakehouse paradigm resolves this tension by layering transactional semantics, metadata management, and open storage formats directly onto cloud object storage. This paper proposes a six-layer Lakehouse reference architecture for financial systems, grounded in the Banking Industry Architecture Network (BIAN) Service Landscape 12.0 service-domain model. Each layer is mapped to the service domains that own its data products, which gives the platform a clear line of accountability. The architecture addresses ingestion and source integration, open table storage using the medallion pattern, governance and compliance, analytical and risk serving, real-time streaming integration, and data product delivery. A structured comparison of Delta Lake, Apache Iceberg, and Apache Hudi across financial workload profiles (batch regulatory reporting, real-time risk scoring, and ML model training) guides open table format selection decisions. Design tensions between real-time processing latency and regulatory auditability are analysed, together with a federated governance model that reconciles domain ownership of data with the enterprise-wide lineage that BCBS 239 requires, and a phased implementation roadmap is provided. The framework provides data architects and platform engineers with a structured foundation for building unified financial data platforms that satisfy BCBS 239, GDPR, SOX, and EU AI Act obligations while delivering the analytical throughput and ML serving capabilities that modern risk management demands.

East African Journal of Information TechnologyVol. 9(3)
Big Data and Business Intelligence
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon