Changes with impact on existing DDL/API functions

Forward Engineering — Snowflake

Snowflake Forward Engineering now offers CREATE OR ALTER as a default action, available in both Full Script and Alter Script; object types that don't support it automatically fall back to CREATE OR REPLACE or DROP + CREATE

Changes with no impact on existing DDL/API functions

AI Copilot

Copilot can now attach an existing column template (e.g. standard audit columns) to one or more tables, or detach a template from a table, instead of doing it manually in the property editor
Copilot now recognizes a project's table templates and can create a new table from an existing template, create a new template, or edit an existing one
The AI Settings → Users tab now notes that only members with an AI role assigned are listed, with a link to the Admin Console for assigning AI roles
(Fix) AI features (Reverse Engineering AI, AI context menu, Documentation AI) are no longer hidden when AI Copilot is enabled — they now follow the single AI Copilot setting

User Connection

User Connection — Multi-Database Support: saved database connections are now available for all six supported databases (Snowflake, BigQuery, Databricks, Redshift, SQL Server, and Azure Synapse). Save your credentials once on the dashboard and reuse them across Direct Connect — Reverse Engineering, Environments, and Compare Revisions — with the connection picker automatically filtered to the current database type.

Databricks

Added support for the VARIANT, GEOGRAPHY, GEOMETRY, and TIMESTAMP_NTZ data types across the UI, Reverse Engineering, Forward Engineering, logical-to-physical conversion, column templates, and copy/paste
Added support for Databricks complex data types: STRUCT (multi-level nesting of any column types), ARRAY (single-type nesting), and MAP (key/value pairs)
Databricks columns now support COLLATE and MASK properties, and TIMESTAMP_NTZ columns are no longer skipped during Reverse Engineering
(Fix) Databricks Bloom filter index names are now preserved on a Forward Engineer → Reverse Engineer round-trip, removing a spurious "Updated" diff

BigQuery

BigQuery primary key columns can now be nullable, matching BigQuery's behavior; the Relationship Panel Editor and validation handle nullable primary keys correctly
(Fix) Saving a BigQuery project no longer fails when an external container is already configured

dbt Manifest Import

(Fix) dbt models are no longer mistakenly tagged as sources when they are also referenced as a source elsewhere in the manifest

Integrations

Slack Integration (MVP): connect your SqlDBM workspace to Slack and receive notifications about project activity directly in your Slack channels

Permissions

(Fix) Governance-role users can again save documentation and model governance fields after a branch is updated from main with new Case Standards settings
(Fix) Project owners on Free and Basic tiers are no longer downgraded to read-only on their own projects after permission changes

Performance

(Fix) Save and merge operations in large projects no longer experience intermittent slowdowns

Dashboard & UI

(Fix) Sort arrows on the dashboard and AI settings tables now point in the correct direction — up for ascending, down for descending