How NASDAQ built a governed intelligence layer with dbt and Databricks

In financial services, a data error doesn’t just mean a wrong number on a dashboard. It can mean a flawed regulatory filing, an incorrect client bill, or a risk team making decisions based on stale information. That’s why the stakes are higher, and why NASDAQ’s approach to data infrastructure is worth a close look.

NASDAQ’s Eclipse Intelligence platform, built on dbt and Databricks, started as an internal solution for the company’s own markets. After a decade of refinement, it’s now offered to external financial market infrastructure (FMI) customers—exchanges, clearinghouses, and central securities depositories. The platform handles up to a trillion messages per day from thousands of sources, and it’s designed to ensure accuracy at every step.

Michael Weiss, AVP of Product at NASDAQ, explains that the platform uses a dbt Mesh structure to enforce consistent data contracts from the start. This means every message on an order chain looks the same, no matter the source. The result? Metrics and KPIs are defined once and reused across the entire organization. If a pipeline test fails, the team would rather delay data delivery than send incorrect information.

Andrea DeSosa from Databricks notes that Unity Catalog provides infrastructure-level governance, tracking every transformation and access point automatically. This is critical for compliance with regulations like SEC Rule 17A and SOC 2, where audit trails must be clear and defensible.

For business teams, the impact is tangible. By putting governed SQL tools in the hands of analysts, NASDAQ reduced time-to-market for new data products by 30-40%. Sales teams can build their own client visuals without engineering support.

And on AI readiness, the semantic layer built with dbt ensures that any AI agent accessing data understands exactly what each metric means. Without it, confident but wrong answers become risk events. NASDAQ is now building agentic workflows on top of this foundation, with a marketplace for sharing governed, verifiable models.

The lesson for business leaders: the same data discipline that satisfies regulators today is what makes AI trustworthy tomorrow. For FMI customers, that foundation is now available in six months, not three years.