Book a demo

Resource Library : WEBINAR

Ontologies Give AI Context. Who Does the Math?

Snowflake, Databricks, Microsoft, and dbt are all investing in ways to give AI richer business context. That’s good news for anyone putting AI on enterprise data. Knowing what each table means and where it lives is the first step toward a right answer.

December 16, 2026 | 2PM ET / 11AM PT (60 Minutes)

Webinar: Ontologies Give AI Context. Who Does the Math?

You still need an engine to do the math. AI has to compute each answer the same way every time, whether the question comes from an agent or a Power BI report. That’s hard when logic is dispersed across BI models, Excel workbooks, and in the heads of subject-matter experts.

In this webinar, AtScale’s Dave Mariani and Mark Palmer join Keith Belanger to show how a universal semantic layer builds on the data platform you’ve already invested in. It turns that platform’s context into answers you can repeat and defend. They’ll draw on AtScale’s BIRD-Interact benchmark work, which compared AI writing SQL straight against database tables with AI working through AtScale’s semantic model and computation engine. AI can be probabilistic in how it finds the data. It has to be deterministic when it does the math.

Key takeaways

  1. Cut down on semantic sprawl. Combine platform semantics and ontologies with dashboards, spreadsheets, schemas, and SME knowledge into one reusable set of business definitions.
  2. Get the same number everywhere. Watch a deterministic engine handle the hard math, from ratios at every level to rolling windows. Given the same definitions, data, filters, and security context, it returns the same answer whether you ask in an agent or in Excel.
  3. Measure how far you can trust AI. Run AI answers as a regression test against gold-standard SQL. Score each metric over time to catch drift when models, prompts, or definitions change, with row- and column-level security applied to every definition.

Register Now