Fivetran + dbt: The Myth of 'Open Data Infrastructure'

Fivetran and dbt Labs just announced an all-stock merger, calling it a milestone for the future of open data infrastructure.
Both companies say this isn't about lock-in — it's about "uniting to set the standard for open data infrastructure."
(dbt Labs announcement) • (Fivetran press release)

Every consolidation starts with talk of openness. But in practice, acquisitions tend to make systems tighter, not freer.

Five years ago, the modern data stack — Fivetran, dbt, Snowflake, and Looker — was the gold standard. Every data team wanted it.
Today, that same stack represents what most are trying to escape: complexity, compounding cost, and slow value realization.

At Definite, we built for the new reality: an AI-native, open-source foundation that delivers full-stack analytics instantly — without lock-in or overhead.

The Familiar Story of "Openness"

Mergers are always sold as "open." The press releases promise interoperability. But behind the scenes, the incentive is control.

When two closed-source SaaS vendors combine, "open infrastructure" becomes a promise that needs constant verification.
Openness is only as strong as the weakest proprietary layer underneath it.

The dbt + Fivetran merger is being positioned as a unification of standards.
In reality, it's a deepening of dependencies — just now under one logo instead of two.

The Frankenstack Problem Isn't Fixed — It's Rebranded

The modern data stack broke because customers had to stitch it together themselves:
Fivetran for ingestion. dbt for transformation. Snowflake for storage. Looker for BI.

Four vendors. Four contracts. Four APIs. And a tangle of billing, permissions, and version drift.

Even if one company owns all those pieces, they're still separate products.
You're still buying, then building. Consolidation doesn't eliminate friction — it just moves it inside the vendor's walls.

The result is the same stack — fewer login screens, higher bills.

That's why we built Definite. Instead of stitching, we start unified.
You sign up and get a complete, AI-native data platform — from ingestion to insight — built on open foundations. It just works.

What Real Openness Looks Like

At Definite, "no lock-in" isn't a tagline — it's architecture.

Everything in the platform is text — SQL and YAML — stored in open formats.
Our semantic layer runs on Cube, which is open source and fully exportable.

No proprietary metadata store. No secret UI logic. No walled garden.

If you want to leave, you take your SQL and YAML with you. That's it.

That's what open infrastructure means in practice — not a rhetorical flourish on a press release.

Foundations vs. "Open-Core"

The distinction between open and open-core matters more than people admit.

DuckLake — our query and compute foundation — is governed independently, not owned by a VC-backed entity.
Cube is community-driven, not monetized through artificial lock-in.

dbt Labs, by contrast, has investors and a growing commercial Cloud product.
Those incentives will always lean toward control.

Open-source branding means nothing if the foundation can't be forked freely.
When the core of your data stack can be revoked by acquisition, you're not building — you're renting.

The "AI-Native" Divide

Snowflake and others are bolting chatbots onto old workflows.
We built Definite so AI is the workflow.

Our agent doesn't just suggest SQL.
It creates pipelines, builds dashboards, relabels semantic models, and manages metadata — descriptions, currencies, formats, and more.

It doesn't assist; it operates.

That's what "AI-native" means: intelligence that orchestrates the stack, not decorates it.

The Economics of the Modern Data Stack Are Collapsing

The modular stack once promised flexibility. Instead, it delivered a tax — multiple vendors billing for what's basically one pipeline.

Teams are paying hundreds per seat for a UI that writes SQL.
That story is getting harder to sell, especially as dbt Core tightens and Fivetran layers on more pricing.

This merger won't fix that. Integration is expensive, and "bundled simplicity" is just a pretext for higher prices.
Costs don't disappear — they migrate.

Because Definite is built entirely on open-source foundations, our costs stay low and margins healthy without gouging customers.
Efficiency isn't a feature. It's the business model.

The Scale Fallacy

Every vendor defends their bloat with "enterprise-grade."
But 99.9% of companies don't operate at Meta scale.

Before DuckLake, we had bottlenecks — now we don't.
We're not built for petabytes. We're built for everyone else: teams that want clean, queryable data, fast answers, and no infrastructure overhead.

Scale is not the default requirement. Insight is.

Customers Are Already Moving

We've seen the shift firsthand.
From startups to public companies with thousands of employees — teams are moving away from the modern data stack because the complexity tax has become unbearable.

They want simplicity, speed to value, and the freedom to actually own their stack.
Those used to be competing goals. Not anymore.

Why the Big Clouds Can't Follow

Snowflake and Databricks can't do what we're doing.
Their entire model depends on supporting legacy customers, legacy architectures, and partner ecosystems.

They can buy components, but they can't start clean.
Their "all-in-one" integrations look coherent on slides but fragmented in practice.

Definite's advantage is focus.
No third-party dependencies. No legacy drag.
No stitched-together stack pretending to be unified.

The Future Isn't Modular or Monolithic — It's Open and Intelligent

This merger isn't a revolution. It's recognition.
The modular modern data stack has reached its limit — too fragmented, too costly, too slow to deliver value to non-technical users.

The next era isn't about connecting more logos.
It's about open foundations and intelligent orchestration — systems that can reason about data, not just move it.

Built on DuckLake. Orchestrated by AI. Unified through Cube.

That's Definite.
No lock-in. No overhead. Instant value.