Changelog

SCIM provisioning is now available. Automatically create users, assign roles, and manage seats in SqlDBM from Okta or Microsoft Entra ID.

From the moment you log in, SqlDBM now follows your preferred theme. No more switching from a bright Dashboard to a dark diagram, so long modeling sessions are easier on the eyes. Your theme is now one setting for the entire app: set it once, and it applies everywhere.

Data stewards, support contacts, access approvers, and security contacts documented in SqlDBM can now go straight to Snowflake as native contacts, so everyone querying the data knows who to ask. Reverse engineering brings existing Snowflake contacts and their assignments into the model, so ownership no longer has to be re-entered by hand. Each ownership field can either generate DDL or stay as documentation only, keeping teams in control of what reaches the warehouse.

SqlDBM now supports Ossie, the open, vendor-neutral semantic model standard developed under Apache by Snowflake, Databricks, dbt, Salesforce/Tableau and others, and you can use it in every project type. You can export your semantic models as Ossie YAML and import them into any project or platform, so you don’t have to rebuild datasets, relationships, metrics, descriptions and AI context when you move between projects, migrate to another warehouse or share with partner tools. Your semantic layer is no longer locked to one vendor’s format, so you can take it anywhere.

Your favorite documentation and metadata fields can now be saved as presets for quick and easy access. Do you maintain dbt properties? Save it as a preset. Frequently access a set of descriptions and governance fields? Save it as a preset. Do you edit Semantic models? You get the idea. Next time you want to edit those fields, just select the preset from the field selection screen on Database Documentation or the Diagram Metadata panel and have them ready for editing in a single click.

Logical projects can now carry their own icon, so it’s clear at a glance which system or data source each model represents. Upload the logo of a source like Salesforce, SAP, or an internal application, and your end-to-end data pipelines become easier to read and navigate. The icon follows the project across the dashboard, recent projects, and global modeling, helping your whole team find the right model faster.

You can now export only the part of your model you need to Excel: a subject area, specific diagrams, or a hand-picked set of objects. Each file contains just that selection, so you don’t have to work through the whole project to find what matters. Related objects are included automatically, so your relationships stay intact when you bring the file back into SqlDBM.

New projects now pick up your Global Standards automatically: check “Automatically map future projects to this standard” in the Map projects dialog, and every new project of that database type inherits your standards from the start. Logical Global Standards projects can now also set data types mapping as a standard, so logical and physical types convert the same way in every mapped project. This completes the latest round of Global Standards updates, all covered in Global Standards.

You can now reverse engineer directly from your dbt manifest using the UI or remotely via API. With the RE method enabled, SqlDBM creates missing tables and views and updates existing ones with their columns, data types, and keys. Keep in mind that a manifest only contains what your dbt project declares in YAML, so models without enforced contracts may come in with missing or untyped columns. Both manifest uploads, the new reverse engineering import and the existing documentation import, now also sync your dbt properties (e.g., tags, meta, model config, and data tests), so the context your team writes in dbt shows up in SqlDBM automatically.

Model governance fields are now part of Global Standards, so you can define your governance fields once and every mapped project uses the same definitions. Teams describe their data consistently across projects, without recreating fields or fixing mismatched names and values by hand. Modelers keep full control of the values they assign to each object and column, while the definitions stay aligned with your organization’s standards.

The POST API can now create revisions directly on Main of a concurrent-working project, so your pipeline can publish changes without a branch-and-merge step. When the approval process is enabled, only approvers can update Main. Every revision-creating endpoint also accepts approval status, comment and flags, so API revisions carry the same metadata as revisions saved in the tool.

Choose your default action, safe scripts, quote options, and per-object overrides in Project Settings, or click Save project settings on the Forward Engineer screen. They’ll pre-fill for everyone on the project, every time.

Global Standards now include color labels, so you can define a shared set of labels once and every mapped project uses the same names and colors, giving you a consistent visual language across all your models. You can also compare revisions of a Global Standards project to see exactly what changed in your standards.

When you push dbt Source and Model YAML to Git, you can now split the output into separate files per source or per model in addition to the single-file option. Switch on Dynamic paths and SqlDBM automatically recreates your dbt project’s folder structure from the paths saved when you imported your manifest, placing each file right next to the models it describes (and falling back to /models/ if an object has no path), with a Preview tab that shows exactly where every file will go before you push.

You can now generate native, platform-ready Metric Views straight from your project, so your measures, joins, and business-friendly definitions go into Databricks without any hand-coding. You can also import your existing metric views to bring those semantic layers into one governed model, where every property is kept on the round trip and your metrics stay in sync from design to deployment.

Model governance fields now travel both ways between SqlDBM and Atlan. Map a schema to your Atlan assets once, then push your field values out to the catalog, or pull the catalog’s values back in and review every proposed change before you accept it. A push that covers more of the model than one request can carry is handed to a background job and carried out in batches, so a large model goes out in one action rather than being split up by hand.

Gain complete, end-to-end visibility into how data models and shared objects connect across every project in your entire organization from a single, centralized view. With the Global Lineage view, teams can eliminate guesswork and prevent costly downstream breaking changes before making schema updates. Search, explore, discover: this unified perspective breaks down organizational silos and gives teams a single pane of glass for object sharing across logical, physical, and business domains.

SqlDBM now offers relationship docking customization for users seeking greater control over diagram layouts. Beyond the default “auto” setting, relationship lines can now be explicitly attached to any cardinal point on a diagram object. This gives you full control over where lines connect and how they flow across your model.

Data Ownership fields bring accountability to Database Documentation: assign a Business Owner, Data Steward, and Technical/Support Contact to any table, pulled from a shared Contacts registry that keeps ownership info consistent across the whole project. Trusted status lets teams flag a table or column as Draft, Verified, or Deprecated, so anyone browsing the model can tell at a glance how much to rely on it. That status now shows right on the diagram, with a color-coded icon on the table header.

Your team’s standards, written once.

An account administrator can now create a skill at organization level, and every project in the account can use it. The governance rules and review processes your team follows live in one place, so Copilot works the same way whichever project you open.

Managing a project’s own skills is no longer the owner’s job alone: anyone with edit rights on the project can now do it.

Copilot can now hand work to specialists, chosen from the new Subagents menu.

• Product guide answers questions about how SqlDBM works from the official help center.
• Planner turns a big request into an ordered plan you can adjust before anything changes.
• Reviewer double-checks finished changes and reports what actually landed.

More of your model, now readable by AI.

A connected assistant now reaches your semantic models, the model governance fields your team maintains, and your dbt YAML, alongside the physical model it already reads.

It can contribute too. Governance changes and dbt manifest imports arrive as a revision for someone to review, exactly like a change made in the app. An assistant sees only what the person whose account is connected can see.

GitHub is now supported through the SqlDBM GitHub App. Install it once in your GitHub org, pick the repositories SqlDBM can reach, and connect your projects.

Colors are now color labels, an enhancement that transforms standard object colors into customizable, named, and described tags to help you better organize, document, and search your projects. All of your existing diagram colors seamlessly transition into color labels with zero disruption to your current project assignments. With this release, you unlock expanded customization options like creating custom colors, added descriptive labels, and an extended curated color palette, giving your team complete flexibility to categorize objects visually.

Model governance fields now travel both ways between SqlDBM and Microsoft Purview. Connect with an Entra app registration, map a collection to your Purview assets once, then push your field values out as business metadata, or pull the catalog’s values back in and review every proposed change before you accept it. A push that covers more of the model than one request can carry is handed to a background job and carried out in batches, so a large model goes out in one action.

Manage all your project-wide configuration from one dedicated settings space, so your team always knows where to look and what applies to everyone. Set your team’s defaults for Reverse Engineering import behavior and for which dbt version, Fusion or Legacy, your project generates by default, so collaborators start from the same baseline. Add a standard header to your generated SQL and dbt code once, and it’s automatically included every time that project’s code is generated, keeping documentation and formatting consistent across your whole team.

SqlDBM now supports materialized views, functions, procedures, sequences, and partitioned tables in PostgreSQL projects. Databases that rely on partitioning for scale, or on materialized views and routines for reporting logic, can now be modeled end to end instead of keeping those objects outside the project. PostgreSQL imports also run up to 15 times faster on the new SQL engine.

Teams can now group diagram objects into named, collapsible containers and turn overwhelming diagrams into a clean, structured view. Collapsing a container hides all the objects and merges all relationship lines into one, dramatically reducing visual noise without losing any of the underlying model detail. This unlocks a new way of working with large models: teams can focus on one domain at a time, share a common diagram layout that everyone sees consistently, and navigate complex data landscapes with confidence.

Push model metadata from SqlDBM straight into Collibra as governed assets, so definitions live in one place instead of being maintained twice. Map SqlDBM fields to attributes that already exist in your Collibra instance, pick the target community and domain, and choose exactly which objects are included. Pushes run as an asynchronous import job you can monitor from the toolbar, and each project can keep several mappings to send different scopes to different domains.

Create reusable AI workflows that capture your organization’s standards, governance rules, and review processes. From documentation and SQL generation to PII detection and model validation, Copilot helps every team follow the same trusted process at scale.

Run all SqlDBM AI through your own model provider and API keys, with support for Anthropic, OpenAI, and Google Gemini. Configure it once in AI Settings by choosing a provider, entering your key, and selecting a model, and Copilot switches over within seconds. Configuration is applied per company and consumes no SqlDBM AI credits, so usage and spend stay under your control. BYOM is enabled for your company by your Customer Success Manager before it can be configured.

Copilot now runs directly on the Semantic Model page, so it always works against the model you have open instead of the project default. Create and switch between semantic models, add tables and views, set business names and descriptions, and configure semantic properties including semantic type, synonyms, sample values, and column exposure. Edits inside a named model override inherited defaults without changing the underlying physical objects, and Copilot reports when an action cannot be completed rather than claiming success.

AI tools connected to SqlDBM can now contribute changes to governed models: create a pending revision, fork a branch, or create a new project from a DDL script. Write access is always explicit — the scope must be approved when you connect, and a read-only token cannot change anything. Governance still applies: agents cannot merge, approve, or delete revisions, and every write uses the caller’s own permissions and lands in the revisions list with their name and timestamp.

Approval Process for concurrent working is here, giving teams a structured, role-based review step before any branch can be merged to Main. This ensures that data model changes are validated by the right people before they reach production. Project owners can enable approvals on any concurrent working project and designate Approvers who review, adjust, and either approve or reject pending merge requests, with automatic notifications keeping all branch members informed at every stage.

Data teams can now programmatically govern their semantic layers using SqlDBM’s public API with GET and POST methods, facilitating effortless integration into CI/CD pipelines and Git-based version control. You can export semantic YAML directly from any project revision to commit it to source control, and then automate updates back to SqlDBM to maintain synchronization. This functionality removes the manual overhead of UI-based management, providing a robust, repeatable workflow for maintaining verified, high-quality definitions at scale.

You can now create a new SqlDBM project directly from an ERwin export, making it easier to bring existing data models into your workflow. This streamlines migration, reduces manual rework, and helps teams start delivering value faster. Simply import your ERwin XML file and continue working in SqlDBM with the target database platform detected automatically.

SqlDBM now generates dbt YAML in Fusion format by default, helping teams adopt the latest dbt workflows with output that is ready to use and free from legacy-format errors and warnings. Teams can continue using the Legacy option for older dbt environments, ensuring existing pipelines remain supported during the transition. The new format is also available in the API as a parameter, with “legacy” set as the default to ensure no disruptions to existing workflows.

You can now organize and categorize your BigQuery resources directly within SqlDBM using native Labels on tables, views, materialized views, and datasets. Lables make it easy to filter, search, and manage large projects by environment, team, domain, or any convention your organization uses. Labels are fully reusable across objects so teams can enforce consistent classification at scale without repetitive manual work.

SqlDBM now supports Databricks tags, letting teams organize, classify, and search their Unity Catalog schemas, tables, views, and columns directly within their data models. You can create reusable tags, assign them in bulk, filter diagrams, and keep everything in sync through forward and reverse engineering. This brings the same consistent tagging experience already available for Snowflake and BigQuery to Databricks users, making it easier to enforce governance standards and find what matters across large catalogs.

The new Global Reference Summary gives you a clear view of which projects are referencing your Global Objects. Available on all Global-enabled projects, the summary table provides sortable details and summary counts make it easier to identify dependencies, coordinate with stakeholders, and avoid unexpected disruptions.

Users now have full control over how primary key columns are positioned within their tables instead of PK columns automatically jumping to the top. Managing and creating primary keys is also more intuitive, with a streamlined constraint menu that lets you add a PK directly in one step, just like any other constraint type. These improvements give you a cleaner, more predictable modeling experience where your table structure stays true to your design intent, both in the diagram and in the generated DDL.

Semantic modeling is now available in Logical projects, helping teams add business meaning to their data earlier in the design process. Users can define semantic defaults once and carry them through forward engineering and Logical-to-Physical import/export workflows, reducing duplicate work and preserving consistency. This continuity makes models easier to understand, maintain, and use for analytics across the development lifecycle.

Semantic Models are now fully supported in Concurrent Working: every team member can create and edit semantic models in their own branch, then review changes in a dedicated Semantic Models tab of the merge screen with a side-by-side, line-highlighted YAML comparison. Your semantic layer gets the same review-and-select merge discipline as SQL objects and diagrams, so multiple modelers can finally work on it in parallel without overwriting each other. Note that a semantic model must be merged together with the objects it references: if its related objects are excluded, the semantic model won’t be merged.

Global References now extend to Logical projects: reference objects across Logical projects or bring logical objects into physical projects or vice versa. Model a shared entity once and reference it everywhere and carry the source project’s icon so your architecture stays readable. Connect your business-level model directly to its physical implementations and give your whole organization one connected picture.

Model, reverse engineer, and forward engineer Teradata in SqlDBM, and bring it into the context layer powering your AI initiatives.

Main is no longer the only option. SqlDBM now lets you merge into any branch you’re a member of! You can also create a new branch from any change branch, so you can build on a teammate’s in-progress work without waiting for it to reach Main.

Reverse engineering is now supported for named semantic models and SqlDBM semantic defaults.

Seamlessly manage virtual relationships directly from your spreadsheets by exporting them alongside physical constraints.

The iframe integration now supports filtering by revision flags, allowing you to display only flagged or approved revisions to your audience.

Branch merging is no longer an all-or-nothing operation at the table level. With our new merge dialog, you can now selectively include or exclude individual column-level changes.

The Get Revision List API endpoint now includes the recently-released Approval and comment fields in addition to all other revision flags.

Revisions now have more metadata fields to record approval status and comments so your team (and your AI agents) have more context on why a change was made and who reviewed it.

Liquibase forward engineering lets teams generate version-controlled changelogs directly from their SqlDBM models, closing the gap between design and automated deployment.

We’ve introduced a new “Whole word” toggle for glossary mapping rules to give you greater precision when enforcing naming standards across your projects.

Column and table templates, Excel export, subtype/supertype hierarchies in logical models, and metadata exploration (what, when, who changed).

Editable view columns, views in database documentation, and more!

All databases with direct connect now supported in User Connections

Allocate AI credits across users from a shared organizational pool, with visibility into assigned, consumed, and remaining balances.

Route account-level notifications to Slack for team visibility while maintaining security alerts via email.

Design a governed, AI-ready semantic layer directly alongside their physical model and forward-engineer it into Snowflake Semantic YAML

The SqlDBM Copilot Context Folder processes uploaded files (Excel, CSV, SQL, Images, PDF) to extract and analyze data models, DDL, documentation, and diagrams.

The SqlDBM Copilot now adds AI-generated summaries to the Compare Revisions and Forward Engineering screens.

Your data model, now readable by AI. The model your team built in SqlDBM becomes the source of truth your AI reads its context from.

Auto-generate Snowflake views and tables from YAML templates, reducing SQL effort from hours to seconds.

Full support for ARRAY and RANGE types (including arrays of structs), property-level Compare that goes beyond plain-text diffing, and editable columns and descriptions on views.

Views now support Virtual Primary Keys and Virtual Business Keys

Forward Engineering now adds quick-filters to instantly isolate dbt sources and models

Snowflake Excel upload support for Dynamic, Hybrid and Iceberg tables

Views and Dynamic Tables now support visible, editable data types as metadata

For automated, system-driven project changes

Redshift news and native syntax support

SqlDBM now supports database.schema.object naming for greater modeling precision.

You can now group projects into dashboard folders with unlimited nesting.

Uploading a dbt manifest now creates virtual dependency relationships from dbt refs.

Connect physical objects back to logical models.

Embed documentation anywhere.

Bridge the gap between your data models and dbt projects with dbt Manifest Import

Model logical joins and data lineage with new virtual relationships

The FE screen is now more streamlined and intuitive

Improved auto-layout, snap-to-grid, alignment guides, and full line reshaping!

Fine-tune word boundaries for control over underscores, numbers, and case changes

FK linking respects column position, and Explorer menus allow Shift+Click multi-select

An AI data modeling assistant that supercharges your productivity

Migrate objects, diagrams, and subject areas from ERwin to SqlDBM

Support for select-all (Ctrl+A) and multiselect (Shift+click) on columns

Flag objects to ignore naming, case, or glossary rules

Iceberg, Dynamic, and Hybrid tables supported in Database Documentation and compare

Fields and Pages editing and merging in Concurrent Working branches

Full native support for all catalog types and properties

Full native support for all properties and indexes

Full native support for Dynamic tables, their logic, and properties

Exclude items from Global Reference at object or schema level

Drag-and-drop functionality for objects in Diagram Explorer

Cancel/reselect revisions mid-compare and schema filtering on Database Documentation

Map predefined standards to your existing projects

Navigable links can now be included alongside your objects

Model Governance fields are now easily accessible directly from the diagram

View existing relationships among global objects without needing to navigate away from the local project

Performance and visual improvements using properties instead of text

Apply rules to physical and/or logical names or set inactive

Case standards for logical names

iFrame embeds for projects in third-party systems

Snowflake views on the diagram with editable columns and parameters

Enhanced foreign key linking experience

Snowflake revamped object parser offers faster performance

Push dbt source & model YAML to git in Forward Engineering

Added support for multiple repos in push-to-git

Simplified UI for Foreign Keys and Relationships