Asia's Real-Time Markets: An LTAP Architecture with Databricks Lakebase and Lakehouse - Hong Kong Databricks FSI Community Day 2026

On the Hong Kong Databricks FSI Community Day 2026, experts gathered aboard a private boat to discuss a novel architecture for managing real-time markets across Asia. The forum, separate from typical conventions, focused on the need for an integrated, governed data plane to handle the complex financial systems of the region.

Proposals presented during the event detailed how Databricks Lakebase and Lakehouse could create a unified architecture, combining transactional processing and analytical computing. This approach aimed to eliminate the latency and duplication issues caused by separating OLTP and OLAP systems. Lakebase, a managed PostgreSQL database, would handle low-latency application state and workflows, while Lakehouse served as a real-time warehouse for millisecond-responsive queries.

The proposed architecture emphasized a single, governed data layer for all market data, ensuring consistency and auditability across the various financial institutions operating in Asia. By integrating Structured Streaming, the system could continuously normalize, enrich, and process data without the latency concerns of traditional micro-batching.

The session also addressed workload isolation, control boundaries, and data residency, ensuring that different entities could access the necessary information while maintaining compliance with regional regulations. The result was a robust, efficient architecture designed to meet the demands of modern, fragmented financial markets.