What Fivetran + dbt Labs brought to Databricks Data + AI Summit (and what you can take home)
Last edited on Jun 23, 2026
If agents are becoming a primary consumer of your data, what does that mean for your data work? It’s a lot to think about, and a lot to talk about, and ask questions about, which is exactly what we heard at the 2026 Databricks Data + AI Summit. Agents are only as reliable as the governed, tested, traceable data beneath them, which is why we showed up in San Francisco last week to show what’s important to us: data infrastructure for agents you trust.
We recently announced that dbt Labs and Fivetran are one company. It makes Databricks Data + AI Summit all the more meaningful to us to show up with our shared mission to build Open Data Infrastructure for the agentic era.
What we shared at Databricks Data + AI Summit
We had a great time at Databricks Data + AI Summit, and if you were there, we hope you had a chance to chat with us. Based on the number of folks that attended our sessions, visited our booth, or attended our events, a lot of you tried. And if you didn’t have a chance, we’d love to talk with you about how you’re approaching the data and agents landscape. Talk to a dbt expert now.
Kicking things off, Fivetran + dbt Labs CEO George Fraser and Fivetran + dbt Labs President Tristan Handy took the stage to share their vision of how AI is reshaping data infrastructure for the next decade. As agents become a primary consumer of data, they query the warehouse continuously, which raises the stakes on cost, observability, and above all, context.
For George and Tristan, the answer to this is Open Data Infrastructure. The context your agents need lives in your data platform. For data teams, that means providing trusted, governed context for AI becomes a core part of the job.
dbt and Fivetran customers sharing their journey
The “BI why” problem
Our customers provide an on-the-ground experience of what it’s like for data teams right now.
Mars Dauer, Senior Director of Enterprise Data and AI at Integral Ad Science, spoke about what he calls the “BI why” problem: when a BI copilot can’t tell you why a number is wrong. It happens because the copilot only sees the top of the stack, and not the business logic underneath that produced the number.
His team’s solution is an AI agent that connects Databricks and dbt through the model context protocol (MCP). The agent draws on three MCP servers, the dbt MCP server for model metadata, lineage, and compiled SQL, a Databricks SQL MCP server for read-only validation governed by Unity Catalog, and a Looker MCP server as the BI on-ramp.
What used to take an analyst an afternoon, now takes minutes, Dauer says, and they are only scratching the surface.
We recently announced dbt Wizard, an agent built for analytics engineers. It works alongside you with deep domain knowledge and a complete understanding of your project and workflows.
Scale data and AI with Fivetran, dbt, and Databricks
Inova Health Chief Data + AI Officer Jon McManus shared how the healthcare provider compressed a 4-year modernization roadmap into 6 months with Fivetran, dbt, and Databricks. His team cut data movement costs by up to 8x with Fivetran. dbt introduced a shared layer for transformation and governance, which enables the team to define consistent metrics and improve trust in data across the enterprise. And Databricks provided a scalable platform to support advanced analytics and AI workloads.
dbt on display at Databricks Data + AI Summit
We’ve announced a lot recently, and it was pure joy giving folks their first look at everything in our booth and sessions and hands-on labs. With Fivetran and dbt Labs now as one company, we’re building the open data foundation for the agentic era so that your business logic and governance travel with you.
We loved talking with folks new to dbt as they discovered the power of the dbt platform. And for current dbt users, we got to show off what’s new:
dbt Wizard is an AI agent purpose-built for analytics engineering to understand your lineage, tests, contracts, and metric definitions.
dbt State skips unchanged models across production and local runs for 30% average warehouse compute savings.
And we announced the first alpha release of dbt Core v2.0, now built on the same foundations as the dbt Fusion engine.
dbt Core v2.0’s key feature developments include:
- Significant parse time improvements
- A tightly-defined language spec that gives anyone integrating with the dbt ecosystem a stable interface to build against.
- New Parquet artifacts as a high-performance alternative to large JSON files that can be directly queried through your agent of choice.
- A completely revamped local documentation experience, powered by those new artifacts and capable of scaling.
- A more streamlined way to build new adapters, powered by Arrow Database Connectivity (ADBC) and the Arrow ecosystem.
- A simplified installation process that removes the need to fight with Python's virtual environments.
The dbt community brings the energy
Whether you stopped by our booth, attended our talks, or joined our Happy Hour by the Bay or executive dinners, we’re grateful to connect with you. The dbt community is doing the most exciting work in data.
If you want to see how the dbt community REALLY shows up, join us at dbt Summit Sept. 15-18 in Las Vegas. dbt Summit is the world’s largest gathering of dbt users. Level up your data and AI work with dbt community members to shape the future of analytics engineering.
Whether it’s at dbt Summit, our webinars, a dbt meetup, or dbt Slack, we’ll see you soon.
Get started in dbt
Join the analytics engineers building data infrastructure that actually scales.
Install dbt Wizard CLI
Get started with an agent purpose-built for analytics engineering. It knows which tool to call, which context to pull, and checks its own work before surfacing anything to you.
