With AI changing the way organizations manage and interact with data, dbt Labs has built a new engine for its data development and transformation software to meet today’s performance and scalability demands.
Dbt Labs this week debuted a new engine for the company’s flagship data development platform that the company said dramatically boosts the system’s performance and scalability – and enhances the data developer experience – for building data pipelines and processing data at scale for AI and analytical applications.
With the new dbt Fusion engine, the company says its platform offers improved developer productivity, higher data velocity and cost savings through more efficient orchestration of data pipelines.
“AI is completely changing the way we interact with data. This is, seriously, the biggest launch in the history of the company and I am super, super excited to see the community response,” dbt Labs founder and CEO Tristan Handy (pictured) said in an interview with CRN.
[Related: Meeting The Data Needs Of The AI World: The 2025 CRN Big Data 100]
Dbt Labs got its start in 2016 as more of a data management consulting firm and it developed its data transformation software as a tool to work with clients.
Handy described Fusion as a complete rebuild of the dbt platform’s engine. “When we originally built dbt, we imagined that maybe a couple dozen companies would use it,” he said. “Now there’s over 60,000 companies using dbt in production,” and he said the original technology was running into performance limitations.
Fusion is written in the Rust programming language, which is seen as superior for building fast and reliable command line interface (CLI) tools – due, in part, to its ability to run multiple computations in parallel. (The original engine was written in Python.)
The new Fusion engine powers the entire dbt platform, including the CLI, dbt Orchestrator, Catalog, Studio and other dbt Labs commercial products, speeding up parse times by a factor of 30x over the original dbt Core, according to the company. That enables faster analytics delivery, lower cloud costs and more trustworthy data pipelines for AI at scale.
Fusion is equipped with native SQL comprehension and other new capabilities that dbt Labs said collectively provide a “best-in-class developer experience.” Those capabilities stem from dbt Labs’ acquisition of SDF Labs, a startup developer of SQL code analyzer tools, in January.
Dbt Labs also unveiled a VS code extension that the company said unlocks broader access to Fusion’s capabilities for local developers. And the company now offers dbt MCP (Model Context Protocol) server, which provides universal connectivity between AI systems and the governed, structured data in the dbt platform.
Fusion is now available for dbt projects on Snowflake with support for Databricks, Google Cloud BigQuery and Amazon Redshift coming soon, according to the company. The VS Code extension is available through the VS Code Marketplace.
Beyond the platform’s core engine, dbt Labs also debuted this week a new suite of AI-enhanced features that the company said provides data analysts working with the Analytics Development Lifecycle with a fast, governed way to explore data and deliver insights within dbt workflows.
Dbt Canvas (now generally available) is a new visual editing environment within the dbt platform, offering a drag-and-drop interface for data model development. Dbt Insights (in preview) is a new AI-powered query tool for quick data analysis and results sharing. And the platform’s dbt Catalog (previously dbt Explorer) has been enhanced for global data asset discovery.
