Databricks expands Lakeflow Designer with visual quality checks and output operators

An October product roundup details no-code cleansing, deduplication, quality checks and expanded outputs in Lakeflow Designer. Databricks identifies app sharing and migration tools as planned rather than shipped.

Databricks outlined Lakeflow Designer with visual quality checks and expanded output operators in a product roundup published on 13 October 2026. The update brings together capabilities delivered over the previous four months, rather than announcing a single October launch: data cleansing, deduplication, validation rules, more flexible joins and additional ways to publish prepared results.

Databricks said Lakeflow Designer had been generally available since June. The visual, no-code tool lets analysts and domain specialists prepare data through operators on a drag-and-drop canvas, with workflows running inside Databricks.

Cleansing and quality checks gain no-code operators

The Prepare operator changes text case, converts data types, replaces values and applies formulas. Unique removes duplicates across entire datasets or selected columns, with ordering controls for keeping the preferred record.

Guardrails adds checks for missing, duplicate or unexpected values. Teams can determine whether validation failures leave passing rows available, block output or stop execution. This makes validation behaviour part of the pipeline configuration instead of relying solely on downstream inspection.

These features arrived in stages. Databricks’ July release notes document Prepare, August notes introduce Unique, and September notes confirm Guardrails. The October post summarises the combined capability rather than dating each operator’s release to 13 October.

Outputs extend to views, tables and files

The Join operator supports multiple input tables and distinguishes matched from unmatched records. Select can pick columns by data type, naming pattern or formula as schemas change. Parameter settings also let source and output configurations vary between scheduled runs.

The Output operator can publish a materialized view in Unity Catalog, merge records into an existing table using selected keys, or append data. It can also write CSV, Excel and JSON files to Unity Catalog volumes. Materialized views refresh when the visual data-preparation workflow runs, according to Databricks.

The company also describes easier navigation for larger canvases, including a searchable table of contents, visible version changes and live collaborator cursors.

Planned features have not shipped

Databricks separately lists future integration with Databricks Apps, full-output data profiling, custom reports and Lakebridge-powered migration tools. Its October article labels those as upcoming, without confirmed release dates.

Lakeflow Designer itself is generally available, and the roundup describes delivered operators and output options. Planned sharing, reporting and migration features should not be treated as already available in customer workspaces.