Embracing the New Era of AI-Driven Analytical Engineering

with dbt Fusion








Scroll to Learn More ↓

Janie Zhan

Melbourne Consultant

Meet Janie Zhan


With a strong foundation in data engineering and analytics, Janie thrives at the intersection of performance and usability. From designing scalable data pipelines to enabling low-code workflows, they bring hands-on experience with tools like dbt, Snowflake, BigQuery, and Databricks. Known for bridging technical depth with business context, they’re passionate about building AI-native stacks that empower both engineers and analysts alike.

Janie Zhan

Melbourne Consultant

Meet Janie Zhan


With a strong foundation in data engineering and analytics, Janie thrives at the intersection of performance and usability. From designing scalable data pipelines to enabling low-code workflows, they bring hands-on experience with tools like dbt, Snowflake, BigQuery, and Databricks. Known for bridging technical depth with business context, they’re passionate about building AI-native stacks that empower both engineers and analysts alike.

The world of data transformation is evolving…Yet again. And once again, dbt is at the forefront. In 2025, dbt Labs launched the Fusion Engine: a reimagined foundation for the next generation of analytics.


This isn’t just a performance boost, it’s a fundamental shift in how engineers and analysts work with data. If you’re a data leader looking to modernise, or an engineer interested in what’s new, here’s what matters most.

The world of data transformation is evolving…Yet again. And once again, dbt is at the forefront. In 2025, dbt Labs launched the Fusion Engine: a reimagined foundation for the next generation of analytics.


This isn’t just a performance boost, it’s a fundamental shift in how engineers and analysts work with data. If you’re a data leader looking to modernise, or an engineer interested in what’s new, here’s what matters most.

Why dbt Fusion Matters

Over the last decade, dbt has become synonymous with modular, scalable transformation who are trusted by 60,000+ companies globally. But in 2025, it's not just about clean SQL and testing. It's about speed, intelligence, and inclusivity across roles.

The Fusion Engine makes dbt faster, smarter, and more collaborative, built from the ground up to support:


  • AI-assisted development with Copilot
  • Real-time, cost-aware coding
  • Visual and low-code workflows for analysts
  • Seamless integration with leading platforms (Snowflake, Databricks, BigQuery, and more)


It’s a major step toward a unified, AI-native data stack.

For Developers: Code Less, Build More

For engineers, this release is all about performance and automation.

The Fusion Engine delivers:

  • 30x faster compile times that drastically reduce wait time and accelerate iteration
  • Real-time impact analysis to instantly visualise the downstream effects of changes
  • Automatic refactoring that previews affected models before committing, so you can avoid breaking things
  • State-aware orchestration that remembers model states and skips unnecessary work, saving on compute costs
  • A more intelligent coding experience that detects errors and autocomplete and suggest smart SQL code

And all of this is built into familiar environments like VS Code, Studio IDE, and dbt Cloud. For developers working on large, complex projects, this removes a lot of the friction and context-switching that slows things down.

For Developers: Code Less, Build More

For engineers, this release is all about performance and automation.

The Fusion Engine delivers:


  • 30x faster compile times that drastically reduce wait time and accelerate iteration
  • Real-time impact analysis to instantly visualise the downstream effects of changes
  • Automatic refactoring that previews affected models before committing, so you can avoid breaking things
  • State-aware orchestration that remembers model states and skips unnecessary work, saving on compute costs
  • A more intelligent coding experience that detects errors and autocomplete and suggest smart SQL code


And all of this is built into familiar environments like VS Code, Studio IDE, and dbt Cloud. For developers working on large, complex projects, this removes a lot of the friction and context-switching that slows things down.

For Analysts: A Seat at the Transformation Table

dbt has traditionally been a developer-first tool. But that’s starting to shift.

New low-code features like dbt Canvas and dbt Insights are giving analysts more autonomy:


  • Canvas offers a drag-and-drop modeling interface, no SQL required
  • Insights enables SQL or natural language queries directly in dbt—along with native visualisation support
  • A fully integrated data catalog brings together lineage, documentation, and governance in one view


This means analysts can participate in the modeling layer, not just consume outputs. That’s a big deal especially in organisations where analysts outnumber engineers but lack the tooling to contribute meaningfully upstream.

Governance and Cost Control, Built In

Fusion also brings smarter controls for security and spend:


  • Cost insights built directly into the UI (available in preview for Snowflake). dbt predicts a cost saving of ~10% due to its state-aware orchestration
  • Semantic layer integration and an expanded catalog for centralised governance
  • SCIM integration for identity management and access control
  • Optimised orchestration to reduce unnecessary processing
  • For enterprise teams trying to balance flexibility with governance, these features offer much-needed transparency and control

Governance and Cost Control, Built In

Fusion also brings smarter controls for security and spend:

  • Cost insights built directly into the UI (available in preview for Snowflake). dbt predicts a cost saving of ~10% due to its state-aware orchestration
  • Semantic layer integration and an expanded catalog for centralised governance
  • SCIM integration for identity management and access control
  • Optimised orchestration to reduce unnecessary processing
  • For enterprise teams trying to balance flexibility with governance, these features offer much-needed transparency and control

The Ecosystem Shift: It’s All Converging

The wider ecosystem is moving in the same direction toward convergence and AI-native tooling:

The Ecosystem Shift: It’s All Converging

The wider ecosystem is moving in the same direction toward convergence and AI-native tooling:

My Take: Why This Update Actually Matters

As a consultant who often switches hats between analyst and engineer, this is the most exciting dbt release I’ve seen. From the engineering side, the speed improvements are game-changing. Compile lag has always been one of those quiet frustrations especially on complex projects. Being able to iterate, test, and deploy faster means more time solving real problems, less time waiting around. And the new Fusion Engine’s understanding of project context makes refactoring far less risky.


On the analyst side, I’ve long felt the pain of being one step removed from the transformation layer. You see things that need changing, but either don’t have the tooling or access to do it. With Canvas and Insights, that’s changing. I can test logic, model transformations, and ask questions in natural language all without needing to open another tool or ask someone else to build it. It’s not perfect yet, but it’s a big step in the right direction.


And finally, the integrations (especially with VS Code) feel seamless. You’re no longer choosing between control and usability. With Fusion, you get both.


While dbt’s not a one-size-fits-all solution, this update makes it a compelling option for a variety of teams and use cases. If you’re currently evaluating whether dbt is right for your business or if you’ve been using it for a while and are unsure about upgrading, this is a great time to explore its potential and assess whether it aligns with your needs. The new features, especially for analysts, make it an exciting tool to consider for both technical and business-driven data initiatives.

As a consultant who often switches hats between analyst and engineer, this is the most exciting dbt release I’ve seen.

From the engineering side, the speed improvements are game-changing. Compile lag has always been one of those quiet frustrations especially on complex projects. Being able to iterate, test, and deploy faster means more time solving real problems, less time waiting around. And the new Fusion Engine’s understanding of project context makes refactoring far less risky.

On the analyst side, I’ve long felt the pain of being one step removed from the transformation layer. You see things that need changing, but either don’t have the tooling or access to do it. With Canvas and Insights, that’s changing. I can test logic, model transformations, and ask questions in natural language all without needing to open another tool or ask someone else to build it. It’s not perfect yet, but it’s a big step in the right direction.

And finally, the integrations (especially with VS Code) feel seamless. You’re no longer choosing between control and usability. With Fusion, you get both.

While dbt’s not a one-size-fits-all solution, this update makes it a compelling option for a variety of teams and use cases. If you’re currently evaluating whether dbt is right for your business or if you’ve been using it for a while and are unsure about upgrading, this is a great time to explore its potential and assess whether it aligns with your needs. The new features, especially for analysts, make it an exciting tool to consider for both technical and business-driven data initiatives.

Ready to Explore?

We're inflight with dbt health checks and quick start sessions for teams who are curious about these new capabilities. Whether you’re just starting out or looking to optimise your setup, now’s a great time to explore what dbt Fusion can offer.


If you’re keen to see how it might work in your environment or just want a walkthrough of the new features, feel free to reach out. We’re always happy to run a demo or chat through what’s possible in plain language, no pressure.

We're inflight with dbt health checks and quick start sessions for teams who are curious about these new capabilities. Whether you’re just starting out or looking to optimise your setup, now’s a great time to explore what dbt Fusion can offer. If you’re keen to see how it might work in your environment or just want a walkthrough of the new features, feel free to reach out. We’re always happy to run a demo or chat through what’s possible in plain language, no pressure.