A calm look at what is quietly arriving on the platform, and what just became real
Databricks moves fast. Really fast.
Most of us feel it but never get a clean view of it. Features arrive quietly in preview, sit there for a while, then one day they are general availability and everyone acts like they were always there.
At BricksNotes we spend a lot of time watching the platform so you do not have to. We keep an eye on what is quietly arriving and what just became real, then we share the parts that actually matter for your day to day work.
This is one of those digests. It is our team view on a set of recent updates, written for Databricks aspirants and working data professionals who want to stay ahead and keep their workflow calm. Read it as a roadmap you can prepare for, not a news bulletin you have to chase.
That is the whole story in two lines. Now let us slow down and understand each one, because the why matters more than the what.
Status: Public Preview
This enables a new Change Data Feed mode for Iceberg writers. Change Data Feed is the idea that a table can tell you what changed, row by row, instead of making you compare full snapshots yourself.
Our take: incremental pipelines live or die on knowing what changed. If the table hands you the changes directly, your downstream jobs get simpler and cheaper, and you spend less time writing fragile comparison logic. Bringing this to Iceberg writers means the open format keeps closing the gap with the Delta experience many of us already rely on.
If incremental thinking is new to you, the incremental processing lesson walks through the idea from the ground up.
Status: Beta
This lets a Databricks App redeploy automatically when you push changes to a connected GitHub repository.
Our take: this is continuous deployment arriving for Databricks Apps. You stop clicking redeploy by hand and let a push do it for you. It is a small feature with a big payoff for productivity, because it removes one of those repetitive manual steps that quietly eat your focus.
If the words CI and CD make you nervous, start with Declarative Automation Bundles for Data Engineers Who Have Never Done CI/CD. It explains the whole idea without assuming a DevOps background.
Status: Beta
This adds workload level query metrics, performance insights, and better run visibility for Lakeflow jobs running on serverless.
Our take: serverless is wonderful until something is slow and you cannot see why. Better metrics turn guessing into reading. When you can see how a run actually behaved, tuning stops being folklore and becomes a habit, and you fix slow jobs in minutes instead of an afternoon.
To build the underlying instinct for reading what a job is doing, see the debugging and monitoring lesson and the partitioning and performance lesson.
Status: Beta
These are open and community built connectors for ingestion sources that the managed Lakeflow Connect connectors do not cover yet.
Our take: every team has that one odd source. Community connectors mean you are not blocked waiting for an official one, and the catalog of what you can pull in grows much faster. For a learner, it is also a great way to see how ingestion really works under the hood.
If you are still deciding how to ingest at all, Lakeflow Connect vs Auto Loader vs COPY INTO is the clearest place to start, along with the data sources lesson.
Status: Beta
This enables materialized views and streaming tables directly from serverless notebooks and jobs.
Our take: materialized views and streaming tables are two of the simplest ways to express incremental, always fresh data. Letting you create them from ordinary serverless notebooks lowers the wall between exploring an idea and shipping it, so you move from prototype to pipeline without changing tools.
To understand the streaming side intuitively, read the streaming lesson, then connect it back to the incremental processing lesson.
Status: Beta
This publishes Unity Catalog tables to Microsoft Fabric as read only mirrored tables.
Our take: data rarely lives in one tool. Mirroring your governed tables into Fabric lets people work where they are comfortable while the source of truth stays in Unity Catalog. For teams, that means fewer copies, fewer arguments about which number is right, and less time spent reconciling.
This is part of a larger pattern of Unity Catalog opening its doors. See Unity Catalog Just Opened Its Doors Wider and the Unity Catalog lesson for the governance foundation.
In this workspace, Iceberg V3 moved to general availability. Managed Iceberg and Foreign Iceberg are also GA.
Iceberg V3 adds deletion vectors, row lineage, and VARIANT support for managed Delta and UniForm and Iceberg scenarios. In plain words, it makes updates and deletes faster, lets you trace where rows came from, and handles semi structured data more naturally.
Our take: this is the kind of change that quietly raises the floor for everyone. The same open table can now do more without bolting on extra tools, which is exactly the sort of detail that shows up in real projects and in certification questions.
If you want the full story of what GA really means here, read Unity Catalog Becomes a Full Iceberg Catalog, and revisit the fundamentals in the Delta Lake lesson.
It is easy to wait until a feature is GA and everyone is talking about it. But the professionals who feel calm during change are usually the ones who saw it coming.
Reading the previews list is like reading the weather before a trip. You are not forced to act. You just get to prepare. You learn the shape of where the platform is heading, so the day a feature lands you already understand the why.
That is the whole spirit of these BricksNotes digests. Not hype. Just a steady, honest look at what is arriving, so you can keep building with a clear head and put your energy where it counts.
This digest is based on the Previews page visible in one Azure Databricks workspace in East US 2. Preview availability varies by workspace, region, and account, so your menu may look different. Treat this as a useful signal from our team, not a universal list.