Starting with this release (7.28), we're rolling out updates in stages over the course of a week to ensure a smooth deployment for all customers. Your app will automatically receive the update as it becomes available. You don't need to take any action. We'll notify you once the release is live in your app. This staged approach helps us monitor each release and address any issues quickly, ensuring the best experience for everyone going forward.

If you have questions, please contact support.

Coalesce Copilot

Coalesce Copilot is an AI assistant that accelerates data pipeline development through natural language interaction. Instead of manually configuring nodes through the UI, simply describe your requirements in plain English or paste existing SQL—Copilot translates your intent into working DAG nodes with proper transformations, columns, and relationships. This intelligent assistant empowers teams to:

• Build and modify nodes through conversational commands.

• Import existing SQL queries into structured Coalesce pipelines.

• Execute multi-step workflows from high-level requirements.

• Receive guidance on best practices and workspace configurations.

By combining the speed of AI-generated scaffolding with the precision of Coalesce's visual interface, Copilot enables both experienced developers and SQL users to build production-ready data transformations faster than ever before.

Enable Coalesce Copilot in your Organization Settings > Preferences.

UI Changes

• Improved Error Messaging for Timed Out Runs — Run timeout errors now say "Run timed out. Restart the run, and if this issue persists contact support." Run details now show whether you executed the run using the Coalesce App or Coa CLI in the new "Executed By" field.

• Readable Add Source Popup After Workspace Creation — The popup that appears after creating a Workspace to guide you on adding sources now has improved text contrast. The message is now easy to read against the blue background.

• Test Failure Indicator in Build Interface — When you run a Node with tests and a test fails with "Continue on Failure" enabled, you'll now see a visual indicator. An orange indicator appears on the Node, and a yellow indicator appears on the failed test child Node. This makes it easier to spot test failures without checking run details.

Mapping Grid

• Prevent Accidental Empty Columns — The Mapping Grid no longer creates empty "nameless" columns when you click or tab through the grid. Previously, clicking in the grid could accidentally generate blank column rows without names or data types. You now need to take an explicit action to create new columns, reducing confusion and metadata clutter.

Bug Fixes

• OAuth Token Storage Issue Resolved — Fixed a scenario where OAuth users with Snowflake connections could be blocked due to a secrets storage issue. The system was unconditionally saving Snowflake tokens in V1 format even when the V2 feature flag was enabled, causing payload size limits to be exceeded in Google Secrets Manager when tokens expired and needed to be reset.

• Full Deploy History Access — Fixed an issue where some users with correct Environment Reader permissions could only view 2-4 recent deploys or runs in the Deploy tab. All users with Environment Reader permissions and above can now see the complete deployment history.

• Node Descriptions Support for Databricks — Fixed an issue where adding a Node description to an existing Databricks Node would cause deployment failures. You can now add or update descriptions on Databricks Nodes without deployment errors.

• Duplicate Column Errors — Fixed a scenario where the deployment process could generate ALTER statements to add columns that already existed, causing production deployments to fail. This issue occurred when certain column metadata changes triggered unnecessary ALTER statements.

Databricks

Incremental Nodes Package

• Version 1.0.0 - The Coalesce Incremental load Node is a versatile Node that allows you to develop and deploy a Stage table/View in Databricks where we can perform incremental load in comparison with a persistent table added on top of it.

Test Utility Package

• Version 1.0.0 - This package is inspired by the Great Expectations package for Python.It will allow you to deploy Great Expectations like tests in your data warehouse.

Lakeflow Declarative Pipelines

• Version 1.0.1 - Parquet file type support added.

Snowflake

Cortex Package

• Version 3.0.1 - Added blank option to all optional dropdown selector configurations, improving usability. Test documentation updated.

• Version 3.0.2 - Fixed Document AI stream name handling and added else blocks to prevent Empty SQL errors. Added default blank for Embedding model in Cortex Search Service.

Dynamic Tables Package

• Version 2.1.4 - Added common column dictionary list macro with logic improvements. Added blank options for Refresh and Initialize configurations.

• Version 2.1.5 - Fix for Missing Handler for Incoming Node type transitions (View/Table → Dynamic Table).

• Version 2.1.6 - Adds Advance Warehouse Selection, Immutability Constraint, and Backfill options across Work/Dimension/Latest Record Version. Updates redeploy/ALTER triggers and simplifies materialization-change steps.

Functional Node Types Package

• Version 4.2.0 - Added new Recursive CTE node type for building recursive queries.

• Version 4.2.2 - Added aggregation transformation support for Recursive CTE. Fixed Empty SQL statement errors.

• Version 4.2.3 - Added Populate on deploy section and Pre-SQL/Post-SQL options for View to Date Dimension Node.

• Version 4.3.0 - New Time Dimension Node for time of day analytics.

Streams and Tasks Package

• Version 2.4.2 - Fixed Empty SQL statement issue. Test documentation updated.

Deferred Merge Package

• Version 2.0.1 - Fixed Empty SQL statement issue.

Data Quality Package

• Version 2.0.1 - Fixed Empty SQL statement issue.

• Version 2.1.0 - Expands DMF node config to include Universal DMFs, Custom DMFs, and Alerting Options; adds static-behavior limitation; updates screenshots and versions table.

Base Node Types Advanced Deploy Package

• Version 2.2.0 - Enhanced delete logic for Fact and Dimension nodes with three distinct methods:

  • Hard Delete (permanent removal)

  • Soft Delete (marks records as isSystemCurrentFlag = 0 while retaining history)

  • No Delete (prevents delete operations to maintain existing data flows).

  • Added Primary Key management configuration for seamless altering and dropping of constraints.

  • Added Merge Exclude List for SCD Type 1, allowing specific columns to be excluded from the update phase of MERGE operations.

• Version 2.2.1 - Fixed Empty SQL statement issue.

• Version 2.3.0 - Added Last Modified comparison to Dimension, Persistent stage, and Fact Nodes.

• Version 2.3.1 - Added Last Modified Comparison CDC options (including Last Modified Column, Treat Null as Current timestamp, and Type 2 Dimension note) across Persistent Stage (table/transient), Dimension (table/transient), and Fact (table/transient).

Base Node Types Package

• Version 1.3.7 - Fixed Empty SQL statement issue.

Iceberg Tables Package

• Version 3.0.3 - Added dictionary functionality for macros in Dynamic Iceberg Dimension and Dynamic Iceberg LRV node types. Fixed Empty SQL errors in Snowpipe Iceberg Table and External Iceberg Table.

External Data Package

• Version 2.0.4 - Added case insensitive toggle to CopyInto, Snowpipe, and External Table node types. Modified run templates for Snowpipe, External Table, and InferSchema to fix Empty SQL errors. Modified External Table to not drop temporary file format in infer process.

• Version 2.0.5 - Adds Snowflake Secrets-based credentials for API node (Bearer/Client ID+Secret) with guidance and images. Adds package tags and updates versions table.

• Version 2.1.1 - Add Git Seet Node. Loads data files directly from Git repositories into Snowflake tables. It connects your version-controlled seed files including CSVs, reference tables, lookup data to your Coalesce pipelines without manual file transfers or separate staging infrastructure.

Interactive Table Package

• Version 1.0.0 - Interactive table deliver low-latency query performance for high-concurrency, interactive workloads.

Incremental Loading

• Version 2.0.1 - Added an else block to the RUN template to handle cases where the materializationType is View.

Dear Coalesce Customers,

We're writing to inform you of an important backend migration that will be completed on or after December 1, 2025 to improve the performance and reliability of the Coalesce platform. We are migrating our database infrastructure from Google Firestore to PostgreSQL, which will enhance our platform's scalability and query performance.

What This Means for You

As part of this migration, some automatic updates will be applied to your workspace configurations to ensure compatibility with our new backend infrastructure. On or after December 1, 2025, when you next load your workspaces, you may notice changes to your YAML files in Git that require your attention.

Additionally, this upgrade will require you to upgrade the version of your Coalesce CLI (coa) to version 7.25 or later to continue to performing operations with our product.

CLI Upgrade Requirements

Due to changes in how we store and retrieve data in our product, we had to make breaking changes to our CLI that require all users to upgrade to version 7.25 or later to allow you to continue operating the CLI. Versions 7.24 and earlier will improperly attempt to read and write to API endpoints that will no longer be supported or used.

How to Upgrade Your CLI:

1. Check your current version:
   coa --version

2. Upgrade to version 7.25 or later:
   npm install -g @coalescesoftware/cli@latest

3. Verify the upgrade:
   coa --version

For detailed instructions, see our documentation.

Please upgrade your CLI by December 1, 2025.

Changes Made to Your Workspaces

The following updates will be automatically applied on or after December 1, 2025 to ensure your workspaces function correctly with our new system:

1. Source Node Cleanup

• Empty sources have been removed - Sources without any configuration or data connections have been deleted from your workspace.

• Null database/schema values corrected - Sources with missing database or schema information have been updated with appropriate default values.

2. Node and Column ID Updates

• Subgraph node IDs standardized - Node identifiers within subgraphs have been updated to maintain consistency.

• Hash column IDs normalized - Column identifiers used in hash operations have been refreshed to ensure proper functionality.

3. Node Configuration Standardization

• SQL node materialization types - SQL nodes without a specified materialization type have been assigned appropriate defaults (typically 'table' or 'view').

• Column name length limits enforced - Columns with names exceeding 256 characters have been truncated or renamed following standard naming conventions.

• Deploy enabled defaults - Nodes with undefined deploy_enabled properties have been set to true by default.

• Column nullable properties - All columns now have explicit nullable values set (true/false) rather than undefined states.

4. Job Cleanup

• Invalid jobs removed - Jobs without proper ID or jobCounter properties have been deleted from your workspace configuration.
  

What You’ll Need To Do After December 1

Immediate Actions Required:

1. Validate Your Workspaces

   • Load each of your Workspaces in Coalesce.

   • Commit any changes triggered by this update to Git via the Git Modal.

   • Verify that all Nodes are functioning as expected.

   • Test critical data pipelines to ensure proper operation.

2. Update Any Hardcoded References

   • If you have any external scripts or references that use the old node IDs or column IDs, these may need to be updated.

   • Review any custom integrations that might reference the modified elements.

Best Practices Moving Forward:

• Column Naming: Ensure column names are under 256 characters.

• Node Configuration: Always specify materialization types for SQL nodes.

• Job Management: Ensure all jobs have proper ID and counter properties.

• Schema Definition: Always specify database and schema values for sources.

Examples of Changes

Here are some examples of the types of changes you might see:

Column Name Length Fix:

# Before (invalid - too long) - name: "this_is_an_extremely_long_column_name_that_exceeds_the_maximum_allowed_length_of_256_characters_and_needs_to_be_truncated_or_renamed_to_comply_with_system_requirements_and_best_practices" 
# After (corrected) - name: "long_column_name_truncated"

Deploy Enabled Standardization:

# Before (undefined) node:   name: my_node   # deploy_enabled property missing
# After (defined)   node:   name: my_node   deploy_enabled: true

Nullable Property Addition:

# Before (missing nullable) columns:   - name: customer_id     dataType: NUMBER
# After (nullable defined) columns:   - name: customer_id     dataType: NUMBER     nullable: false 

Support and Documentation

Getting Help

If you encounter any issues or have questions about these changes:

1. Check the Problem Scanner - Use the built-in Problem Scanner in the Coalesce interface to identify any workspace issues

2. Contact Support - Reach out to your Customer Success Manager or submit a support ticket

Support Contact Information:

• Email: support@coalesce.io

Questions or Concerns?

We understand that any change to your development environment can be concerning. Our Engineering and Support teams have thoroughly tested this migration process, and we're confident that these changes will improve your experience with Coalesce.

If you have any questions, concerns, or encounter any issues, please don't hesitate to reach out to us. We're here to ensure a smooth transition.

Additional Resources:

• Coalesce Documentation

• Git Integration Guide

• Coalesce Support

We’ve introduced caching to make deployment plans faster. The new Caching Deployment Plan feature reduces plan generation time by skipping SQL re-renders for Nodes that haven’t changed.

The feature can be enabled in two ways:

• CLI flag: coa plan --enableCache

• UI toggle: Available in the "Parameters" step of the deployment wizard.

Learn more about using deployment caching in our documentation.

Node.js Version Support Has Been Updated for the Coalesce CLI

With version 7.24, we have updated the supported Node.js versions for the coa CLI to ensure you are working in a stable and secure environment.

What has changed?

The coa CLI now supports the following Node.js versions:

• 20

• 22

• 24

Support for Node.js version 18 has been removed.

Why We Made This Change

We removed support for Node.js version 18 because it is approaching its end-of-life. We have added support for version 24, a Long-Term Support (LTS) version, to ensure continued stability and security.

Action Required

If you are running Node.js version 18, you must upgrade to a supported version (20, 22, or 24) to continue using the coaCLI. If you attempt to use the CLI with Node.js 18, it will exit with an error message stating that the version is not supported.

Metadata Updates

Coalesce has introduced updates to improve metadata integrity across Nodes. In this release, null values in Node metadata are automatically removed.

When you commit changes, you may notice updates to Nodes that you haven’t directly modified. This is expected, since null fields are being removed. You should commit these changes and continue with your workflow.

What to Expect

• Git Modal: You may see diffs in the git modal where null values are removed.

No additional action is required. Commit the changes when they appear to keep your project in sync.

Updates

We’ve made a series of improvements focusing on UI consistency and run visibility.

UI Changes

• Improved Display of Long Storage Location Names - Long storage location names in your build settings are now shortened to fit the screen. You can hover over the shortened name to see the full name in a tooltip.

• Updated the "Re-Sync Source Columns" Window - We've updated the style of the "Re-sync Source Columns" window to match the rest of the application. It now provides a clearer message when there are no changes to sync.

• Improved the "Delete Column" Confirmation - When you delete one or more columns, the confirmation window now clearly lists the columns you are about to delete. This helps you confirm your changes before they are made.

• Updated Git View After Commit – The Git window has been improved so that after a successful commit and push, the view now clears automatically to show "No Files To View" This provides clearer feedback and ensures the view reflects the current state of your repository."

Run Updates

• Added a "Rendering" Status for Runs - When you start a run, you will now see a "Rendering" status indicator. This provides better visibility into the run process, especially for large jobs that take time to render SQL before execution.

• Added Tooltips to Run Details Page - On the Deploy and Refresh Runs page, you can now hover over truncated node names and error messages to see the full text in a tooltip.

• Redesigned the Run Stage Details Window - We've redesigned the window that shows the details for a run stage. The new design makes it easier to read long SQL queries and error messages, with a wider layout and resizable panels.

Git Updates

• Better Error Handling for Existing Git Branches - When creating a new workspace, you will now see a clear error message if the Git branch name you entered already exists, allowing you to choose a unique name.

Package Updates

Base Node Types – Advanced Deploy

• Added comprehensive metadata update checks for Advanced Deployment Templates.

• Fixed block comment issue for metadata update.

Streams and Tasks

• Delta Stream Merge updates:

  • Dual merge logic implemented.

  • Redeployment behavior added for column schema changes.

  • External Table and External Iceberg Table supported as source objects.

  • Deployment phasing introduced.

• Create Templates:

  • Added comprehensive metadata update checks.

  • Fixed comment block parsing.

  • Fixed default scheduler in node definition.

Running Nodes created with any version 2.2.0 and below without adding table key and partition key config options can cause rendering errors.

Create or Alter

• Added comprehensive metadata update checks for Create Templates.

• Fixed comment block parsing.

• Fixed view issues caused by join changes.

External Table

• Partitioning logic updated.

A default column PARTITION_COLUMN is added with the transformation:

split_part(metadata$filename, '/', 2) 

You can rename this column or adjust the transformation. You can also use another column with the right transformation for partitioning.

Bug Fixes

• Prevented Deployments with Empty Column Names - To prevent errors, you can no longer deploy Nodes that have columns without names. The application will now show an error message, "Empty Column Names, during the deploy plan, prompting you to name all columns before deploying.

• Fixed Drag-and-Drop for Jobs and Subgraphs - We fixed an issue where dragging and dropping nodes into a second or third Job or Subgraph would not work. You can now reliably add nodes to any Job or Subgraph using drag-and-drop.

• Fixed Workspace Loading for Nodes with Invalid Configurations - We resolved an issue that prevented workspaces from loading if a node had an invalid configuration value. The application now automatically corrects these values to ensure your workspace always loads correctly.

• Fixed Resizing of Include/Exclude Filters - We fixed a bug where the include and exclude filter boxes in the browser would shrink unexpectedly after being resized. Both filter boxes now resize correctly.

Coalesce has introduced updates to improve metadata integrity across Nodes. Starting in the 7.24 release, null values in Node metadata are automatically removed.

When you commit changes, you may notice updates to Nodes that you haven’t directly modified. This is expected, since null fields are being removed. You should commit these changes and continue with your workflow.

What to Expect

• Git Modal: You may see diffs in the git modal where null values are removed.

No additional action is required. Commit the changes when they appear to keep your project in sync.

Node.js Version Support Has Been Updated for the Coalesce CLI

With version 7.24, we have updated the supported Node.js versions for the coa CLI to ensure you are working in a stable and secure environment.

What has changed?

The coa CLI now supports the following Node.js versions:

• 20

• 22

• 24

⚠️ Support for Node.js version 18 has been removed.

Why we made this change

We removed support for Node.js version 18 because it is approaching its end-of-life. 5We have added support for version 24, a Long-Term Support (LTS) version, to ensure continued stability and security. 6

Action Required

If you are running Node.js version 18, you must upgrade to a supported version to continue using the coa CLI. After updating to release 7.24, running the CLI with Node.js version 18 will cause the CLI to exit with the following error message:

Invalid node version 18 not supported

Updates

We’ve made a series of improvements focusing on UI consistency, run visibility, and Git.

UI Changes

• Improved Display of Long Storage Location Names - Long storage location names in your build settings are now shortened to fit the screen. You can hover over the shortened name to see the full name in a tooltip.

• Updated the "Re-Sync Source Columns" Window - We've updated the style of the "Re-sync Source Columns" window to match the rest of the application. It now provides a clearer message when there are no changes to sync.

• Improved the "Delete Column" Confirmation - When you delete one or more columns, the confirmation window now clearly lists the columns you are about to delete. This helps you confirm your changes before they are made.

• Updated Git View After Commit – The Git window has been improved so that after a successful commit and push, the view now clears automatically to show "No Files To View" This provides clearer feedback and ensures the view reflects the current state of your repository."

Run Updates

• Added a "Rendering" Status for Runs - When you start a run, you will now see a "Rendering" status indicator. This provides better visibility into the run process, especially for large jobs that take time to render SQL before execution.

• Added Tooltips to Run Details Page - On the Deploy and Refresh Runs page, you can now hover over truncated node names and error messages to see the full text in a tooltip.

• Redesigned the Run Stage Details Window - We've redesigned the window that shows the details for a run stage. The new design makes it easier to read long SQL queries and error messages, with a wider layout and resizable panels.

Git Updates

• Better Error Handling for Existing Git Branches - When creating a new workspace, you will now see a clear error message if the Git branch name you entered already exists, allowing you to choose a unique name.

Bug Fixes

• Prevented Deployments with Empty Column Names - To prevent errors, you can no longer deploy Nodes that have columns without names. The application will now show an error message, "Empty Column Names, during the deploy plan, prompting you to name all columns before deploying.

• Fixed Drag-and-Drop for Jobs and Subgraphs - We fixed an issue where dragging and dropping nodes into a second or third Job or Subgraph would not work. You can now reliably add nodes to any Job or Subgraph using drag-and-drop.

• Fixed Workspace Loading for Nodes with Invalid Configurations - We resolved an issue that prevented workspaces from loading if a node had an invalid configuration value. The application now automatically corrects these values to ensure your workspace always loads correctly.

• Fixed Resizing of Include/Exclude Filters - We fixed a bug where the include and exclude filter boxes in the browser would shrink unexpectedly after being resized. Both filter boxes now resize correctly.

Package Updates

Base Node Types – Advanced Deploy

• Added comprehensive metadata update checks for Advanced Deployment Templates.

• Fixed block comment issue for metadata update.

Streams and Tasks

• Delta Stream Merge updates:

  • Dual merge logic implemented.

  • Redeployment behavior added for column schema changes.

  • External Table and External Iceberg Table supported as source objects.

  • Deployment phasing introduced.

• Create Templates:

  • Added comprehensive metadata update checks.

  • Fixed comment block parsing.

  • Fixed default scheduler in node definition.

Running Nodes created with any version 2.2.0 and below without adding table key and partition key config options can cause rendering errors.

Create or Alter

• Added comprehensive metadata update checks for Create Templates.

• Fixed comment block parsing.

• Fixed view issues caused by join changes.

External Table

• Partitioning logic updated.

• A default column PARTITION_COLUMN is added with the transformation:

  split_part(metadata$filename, '/', 2)

You can rename this column or adjust the transformation. You can also use another column with the right transformation for partitioning.

Update to Supported Node.js Versions for the Coalesce CLI

Starting with the 7.24 release, we are updating the supported Node.js versions for the coa CLI to ensure you are working in a stable and secure environment.

What's Changing?

The coa CLI will now enforce the use of the following Node.js versions:

• 20

• 22

• 24

⚠️ Support for Node.js version 18 will be removed.

Why We're Making This Change

We're removing support for version 18 because it reaches its end-of-life in April 2025. Support for version 24, a Long-Term Support (LTS) version, has been added to ensure continued stability and security.

Action Required

If you are running Node.js version 18, you must upgrade to a supported version to continue using the coa CLI. After updating to release 7.24, running the CLI with Node.js version 18 will cause the CLI to exit with the following error message:

Invalid node version 18 not supported

Updates

• Column lineage - We’ve improved the process by adding in app guidance and improving the visual indicators making it easier to see the propagation before commiting.

• Snowflake Authentication - The Problem Scanner now alerts you when Snowflake authentication is missing.

• Subgraphs and Jobs

  • The process for creating a Job or Subgraph is now easier. Enter a name in the modal, click Create, and you’ll be taken straight to your new Job or Subgraph.

  • We alert you in the Problem Scanner for duplicate Subgraph or Job names.

  • Previously, when a was Node was added or removed from a Subgraph it would be done silently. Now we let you know it was successful or failed along with either the Node name or number of Nodes.

Bug Fixes

• When editing user roles, the role summary shows in the dropdown correctly.

• Testing a Databricks SQL Warehouse connection will no longer fail with a "Missing token" error when a valid token has been entered.

• Service accounts with Environment Admin permissions for Snowflake environments will no longer receive an "Invalid Connection Information" error on API calls.

• Deploying a Storage Mapping change will no longer result in a "No schema found" error if the old schema or database was deleted directly from the data warehouse before deployment.

• Deleting a source from a column now works consistently when using a trackpad or a mouse.

• Sorting and filtering by Node Status or Stage Status on the deployment runs page will no longer revert or disappear on runs with a large number of objects.

• Deleting a Source Node and creating a new table Node with the same name and location in a single deployment will no longer replace the existing table and cause data loss.

Understanding how transforms connect back to their source columns is now clearer and more intuitive. This update reduces confusion in the mapping grid by showing source details alongside transforms, giving you a complete and accurate view of column lineage.

Key Improvements

• Clear Source Visibility: When a transformation is applied, the corresponding column source details are displayed in a read-only, grayed-out format.

• Hover for Full Details: Hovering over the source reveals:

  • The original table or alias used

  • All columns included in the transform

  • Automatic resolution of unqualified column names to their source tables

• Hash Column Transparency: Hash columns now expand to show the list of underlying source columns.

• Better Alias Handling: Joins using aliases now display clearer lineage, improving traceability.

Deployments should be clear, organized, and easy to follow. We’ve redesigned the Coalesce deployment experience to make it cleaner, more intuitive, and more informative. The new workflow improves everything from commit selection to error handling.

Key Improvements

• Streamlined Interface: A cleaner deploy view with more space to review commits and deployment details before execution.

• Tabbed Review Experience: Instead of collapsible menus, deployment changes are organized into clear tabs for Nodes, Subgraphs, Jobs, Mappings, Node Types, Macros, and Packages.

• Detailed SQL Visibility: Quickly see what’s changing with at-a-glance summaries, then drill into step-by-step SQL changes for each affected element.

• Improved Validation & Error Handling:

• Parameter validation, for example, JSON format checks, surfaces issues before deployment begins.

• Deployment-time issues, like missing or invalid Storage Locations, now have clearer error messages with actionable guidance to resolve them.