Databricks Announces LTAP Architecture Combining Transactional and Analytical Processing on Apache Iceberg

Databricks CEO Ali Ghodsi unveiled LTAP at the Data + AI Summit in San Francisco. The architecture lets transactional and analytical engines operate directly on a single copy of data stored in Apache Iceberg. LTAP stands for Lake Transactional/Analytical Processing.

It is built on Lakebase, the serverless PostgreSQL database Databricks obtained through its roughly $1 billion Neon acquisition. Lakebase reached general availability on AWS in February. It grew revenue twice as fast as the Lakehouse business in its first six months and now handles millions of database launches a day.

Customers include Block, Superhuman, and Zillow. Ghodsi said the new system addresses a 40-year separation between OLTP and OLAP databases. "For forty years we’ve lived with a separation between OLTP and OLAP because the workloads were genuinely different," he stated.

Databricks introduced synchronization capabilities in 2025. LTAP replaces that approach by eliminating the need for separate copies and pipelines between operational and analytical systems. The company also unveiled Lakehouse//RT, a real-time analytics engine.

Its execution engine, Reyden, sustains sub-100-millisecond latency at up to 12,000 queries per second. Databricks says roughly 80% of databases on its platform are now created by agents rather than humans. Ghodsi said this shift requires infrastructure that supports instant access to fresh data without maintaining multiple copies.

4 billion in annualized revenue. Ghodsi stated the firm is not fundraising and does not have term sheets from anyone. Databricks coined the term "lakehouse" in 2020.