Hong Kong Mortgage Corporation Limited (HKMC) plays a central role in Hong Kong's housing finance ecosystem, providing the infrastructure that helps residents access stable, sustainable mortgage solutions. In June 2024, the organization launched its Data Analytics Platform (DAP), unifying loan application and web analytics data. As demand for richer insights grew, HKMC initiated Phase 2 with a clear ambition: to strengthen the DAP’s sustainability, scalability, and adaptability so it could keep pace with the evolving business, risk, and regulatory needs. LPS was engaged to support this evolution.
FINANCIAL SERVICES
Challenges
HKMC's DAP faced several structural limitations. Without an automated backup or recovery process, responding to incidents, investigations, and data corrections all required manual effort that added friction and risk to day-to-day operations. Data ingested from multiple sources lacked standardization, forcing analysts to spend more time reconciling discrepancies than generating insights.
Compounding this, data quality issues often went undetected, with limited notifications and insufficient visibility into quality metrics, making it difficult for teams to catch and address problems early. Manual ETL testing further increased the risk of errors propagating downstream, undermining the reliability of the reports and dashboards that business stakeholders depend on. Taken together, these gaps pointed to a platform that needed stronger foundations to scale with confidence.
Solutions
LPS’s approach was deliberate and precise: rather than replacing what HKMC had built, the focus was on enhancing and extending it. Working within the existing Azure Databricks environment, LPS designed and implemented an automated backup and recovery mechanism that protected data assets continuously — without disrupting live ETL workflows — while also reviewing and enhancing existing pipelines to improve processing capacity and runtime. With resilience established, LPS turned to the challenge of data consistency, introducing a centralized standardization framework implemented via standardized functions and standardized gold views. By applying uniform mappings, regex-based transformations, and outlier detection across standardized gold views, fragmented and inconsistent records were transformed into a coherent, trustworthy data layer.
To sustain data quality, LPS deployed an automated testing framework with a configurable rules engine and comprehensive test case library — validating data correctness across every medallion layer and alerting users immediately when checks failed. For cases that fell outside the scope of automation, a data patching script provided a governed, controlled mechanism for post-standardization corrections. Tying the entire solution together, LPS aligned data quality rules with the testing framework and surfaced violations through a dedicated Power BI dashboard, giving HKMC's teams clear, real-time visibility into platform health and the actionable insight needed to resolve issues with confidence.
Value created
The upgraded platform empowers a more reliable, efficient, and trustworthy way of working with data for HKMC. By automating testing processes that were previously manual, teams regained meaningful time to focus on analysis and insight generation rather than validation overhead. Standardization and continuous quality assurance elevated confidence in the data underpinning HKMC's reports and analytics — ensuring that when stakeholders act on what they see, they can do so with certainty.
With built-in backup and recovery, HKMC is better equipped to handle audits, investigations, and unexpected incidents with resilience and composure. The DAP has evolved from a reporting tool into a scalable, future-ready data backbone — one that HKMC can confidently build on with further data source integrations as its mission and ambitions continue to grow.
DOWNLOAD PDF
