Databricks defines the Genie Ontology the same way on all three clouds: “the unified context layer that gives Genie One a business-aware map of your organization.” Accurate, and almost useless if you are deciding whether it matters to your company. Here is the plain version.

When someone asks a question of your company data in ordinary English, the system has to work out what the words mean. Which table holds revenue. Whether “active user” counts a person once or once per device. Which of the four dashboards named “Sales” is the one people trust. The Genie Ontology is where those answers live. It is the difference between a system that can read your data and one that understands your business.

The part worth a leader’s attention is not the technology. Half of this layer is defined on purpose. The other half assembles itself, from whatever your teams have already built. Click through the figure below to see both halves and what happens when they disagree.

Inside the Genie Ontology Four stages, from what your team defines to the answer a person receives. Every element below is named in the Databricks documentation.
Stage 1 · Unity Catalog semantics

The half somebody has to decide

Databricks calls this “modeled context that you define, govern, and certify.” It is four tools, and they sit at four different levels of maturity, which is worth knowing before you plan around them.

Metric viewsReusable definitions of a business KPI, governed with normal Unity Catalog permissions. The docs page carries no release-status label.
DomainsPublic Preview Groups data by business purpose, so people browse by what something is for rather than by where it is stored.
PagesBeta Governed definitions of a business concept, written for both people and the system to treat as authoritative.
CertificationMarks an asset as trusted or outdated. The docs say it steers Genie One toward vouched-for assets. No release-status label either way.
This is the half you can direct, and the only half you can hold anyone accountable for. None of these tools fill themselves in.
Stage 2 · Inferred context

The half that assembles itself

Databricks describes this as context the system “automatically extracts and maintains” from work your company has already done. The documented sources are metric views, dashboards, SQL queries, and Genie Agents. Each piece is called a snippet, and the docs name three kinds.

Metric definitionsThe documented example: an “active user” is a distinct user, deduplicated across all platforms.
Authoritative sourcesThe documented example: revenue questions should be answered using the curated Finance Genie Agent.
Business rulesThe documented example: a “qualified lead” only counts once a demo is booked.
Where they come fromMetric views, dashboards, SQL queries, and Genie Agents. Chat can also reach external tools, but those are documented separately.
Read that list of examples again. Those are the definitions that currently live in analysts’ heads and in footnotes on slides. This is where a company’s quiet disagreements about what counts as revenue finally surface.
Stage 3 · The authority score

What happens when two definitions disagree

They are not merged and nobody is asked to choose. Each snippet carries an authority score, and the ranking picks a winner at the moment the question is asked. The documentation names three inputs.

Where it came fromThe asset that produced the snippet
How often it is usedUsage across your workspace
How fresh it isRecency of the underlying asset

Databricks then ranks the most relevant snippets and, in the docs’ words, “resolves conflicts.” Human-modeled context is prioritized over inferred context, though the wording is prioritization rather than a guaranteed override.

Worth knowing: the announcement blog describes five ranking factors under a PageRank comparison, including the authority of a definition’s author. The product documentation names only three. We could not confirm which reflects shipping behaviour, and the algorithm is not published or auditable by customers.
Stage 4 · The answer

Permissions still decide what anyone sees

Snippets are gated by Unity Catalog permissions. Databricks states that Genie One answers “using only the sources you are permitted to see.” The ontology changes how well a question is understood. It does not widen anyone’s access.

QuestionAsked in ordinary English
Ranked contextOnly snippets this person may see
AnswerWith citations back to the source
Databricks also says the narrowing cuts latency, because there is less to crawl and query. Faster and better grounded come from the same mechanism.

What this actually changes for a business

The honest summary is that the Genie Ontology moves an argument, rather than settling it. Companies have always disagreed about what a metric means. Until now that disagreement was survivable, because each team ran its own report and the numbers only collided in a meeting once a quarter.

Put a system in front of everyone that answers in plain English, and the collision happens every day, in public, in front of whoever asked. So the work is not technical. Someone has to decide what “revenue” means and write it down where the system can read it. That is a leadership task wearing a data-tooling costume.

Four things the documentation does not promise

We read the primary sources rather than the launch material, and the gaps are as informative as the features.

  • It is not generally available. The Genie Ontology carries a Public Preview label on all three clouds, with no announced date for general availability. Preview means no service commitment and terms that can change.
  • You cannot edit an individual inferred snippet. No documented interface exists for reviewing, approving, downranking, or deleting one. You steer it indirectly, by curating Pages, metric views, domains, and certification.
  • There is no documented off switch. It has been on by default since August 6, 2026. The same documentation page describes preview toggles for other features, so the absence here is telling rather than accidental.
  • The accuracy numbers are the vendor’s own. Databricks published a benchmark alongside the launch, footnoted as an internal test of 28 questions. No independent measurement of the ontology’s contribution exists. Treat the figures as a vendor claim, because that is what they are.

What we would do about it this quarter

Nothing here requires a project. It requires a decision and an owner.

  1. Pick the five metrics your leadership actually argues about. Revenue, active user, qualified lead, churn, and whatever your industry adds. These are the ones worth defining first.
  2. Give each one a written definition and a name. Pages and metric views exist for this. The tool matters less than the fact that a person decided.
  3. Name an owner per definition. Not a committee. When the definition needs to change, someone has to be allowed to change it.
  4. Assume it is already running. The inferred half has been building itself from your dashboards and queries since August. The question is not whether to start, it is whether anyone is watching what it concluded.

Here is the uncomfortable version. The system is already forming a view of what your business terms mean, assembled from whatever your teams happened to build. If nobody writes the real definitions down, that inferred view becomes the definition by default.

Where these facts come from

Every claim above traces to Databricks documentation or an official post. The ontology has no dedicated documentation page. It is a section inside the “Chat in Genie One” page, which is itself worth knowing when you judge how settled the feature is.

  • Chat in Genie One, the ontology section, where the definition, the two halves, the snippet types, and the authority score are all documented.
  • Unity Catalog semantics, covering metric views, domains, pages, and certification.
  • The announcement post, which is where the PageRank comparison and the benchmark figures appear.

One correction to our own earlier reporting. In our July newsletter we wrote that steward certification feeds into the Genie Ontology. The documentation does not support that as written. Certification is documented as steering Genie One toward trusted assets. The sentence that names the ontology directly is about domains, and it appears on the Databricks blog rather than in the docs.