Every medium-to-large business today has a data plumbing problem. The system that runs day-to-day operations, your orders, your claims, your transactions, stores data in one place. The system that runs your reports and AI tools reads data from another place. Keeping those two places in sync requires a set of constantly running pipelines that extract data, reformat it, and push it across. It is fragile work. Pipeline maintenance alone can run a small data team between ten and fifteen thousand dollars a month, and the actual spend is routinely two to four times higher than companies plan for because of hidden costs like failed jobs, schema mismatches, and compute overruns.
Databricks has now built a way around that. Its database product, Lakebase, now includes a feature called Native Lakehouse Sync. The idea is simple: because Lakebase stores operational data in the same underlying cloud storage as the analytics platform, data does not need to travel anywhere. A single toggle in the settings propagates every change automatically, within about a minute, at no extra cost.
The deeper move here is architectural. Cloud providers like AWS, Google, and Microsoft have all offered versions of this promise before, under names like zero-ETL. But those offerings still route data between two separate systems that happen to be owned by the same vendor. Databricks is making a different claim: that the operational database and the analytics platform share the exact same physical storage, so sync is not really sync at all. It is more like one system becoming aware of its own data.
This matters for three practical reasons. First, when the schema of a database table changes, a traditional pipeline will often break and require manual intervention. With Native Lakehouse Sync, Databricks says schema changes flow automatically. Second, every change is recorded as history, which means compliance and audit trails come for free without a separate logging setup. Third, because the data lands directly in the analytics platform as a governed, queryable table, any AI model or report running in that environment sees up-to-date information without a delay.
The competitive picture is also important context. In May 2025, Databricks spent one billion dollars acquiring Neon, a startup that provided the serverless Postgres technology now underlying Lakebase. Less than three weeks later, Snowflake announced it would spend $250 million to acquire Crunchy Data and build its own integrated Postgres offering. Both of these companies reached the same conclusion at roughly the same time: the next platform war is over who owns the operational database layer, not just the analytics layer.
Adoption has been notably fast. Since Lakebase launched in June 2025, Databricks reported that uptake grew at more than twice the rate of its earlier data warehousing product, with thousands of companies running production workloads on it. EasyJet, for example, used Lakebase to replace a decade-old database environment, cutting development cycles from nine months to four. A shipping company called Hafnia reduced its app delivery time from two months to five days.
For companies not yet on Databricks, the relevant signal is this: the tooling industry that currently charges separately for pipeline management, change data capture, and data sync is being absorbed into platform products. The standalone pipeline vendors, the consulting firms who build and maintain these connections, and the internal data engineering teams whose primary job is keeping pipelines alive, all face a narrower market over the next three to five years. The pressure is not theoretical. It is already showing up in the enterprise deals being signed today.