What is an enterprise context layer?

Summary

• An enterprise context layer gives AI agents the business meaning behind your data: what terms and metrics mean, which sources are authoritative, and which rules and permissions apply.
• Databricks describes Genie Ontology as "the unified context layer that gives the Genie family of AI experiences a business-aware map of your organization". It is in Public Preview.
• It combines Unity Catalog semantics that you define, govern and certify with context that Genie infers from your assets and usage.
• Each inferred snippet gets an authority score, and Genie answers using only the sources you are permitted to see.
• Genie One and Genie Code search the same ontology.

What is an enterprise context layer?

An enterprise context layer is the business context that AI agents draw on to interpret an organization's data: what its terms and metrics mean, which sources are authoritative, and which rules and permissions apply. A Databricks blog post on data ontologies draws the line this way: "A schema describes how data is structured. An ontology describes what that data means in the context of the business."
The post gives a revenue example: "A schema might tell an agent that a table contains net_rev, gross_rev and recog_rev. An ontology can help it understand which definition of revenue applies to the question, which source Finance considers authoritative, how the calculation is normally performed and whether the person asking is even permitted to access it."

How does Databricks define it?

Databricks documents its version as Genie Ontology: "the unified context layer that gives the Genie family of AI experiences a business-aware map of your organization." The docs label the feature Public Preview. The ontology brings together two categories of context: Unity Catalog semantics, which is "Modeled context that you define, govern, and certify through features like metric views, domains, and Pages", and inferred context, which is "A map of snippets that Genie automatically extracts and maintains from your existing assets and usage, including metric views, dashboards, SQL queries, and Genie Agents."

Part	Kind	What the docs say
Metric views	Modeled	"Reusable SQL objects that define and govern business KPIs"
Pages (Beta)	Modeled	"a governed, authoritative definition of a business concept, such as a critical term, entity, or acronym"
Metric definitions	Inferred	For example, "An 'active user' is a distinct user, deduplicated across all platforms."
Authoritative sources	Inferred	For example, "Revenue questions should be answered using the curated Finance Genie Agent."
Business rules	Inferred	For example, "A 'qualified lead' only counts once a demo is booked."

How does Genie choose which context to use?

"Each snippet gets an authority score based on where it was generated from, how often it is used, and how fresh it is." When you ask a question, Genie "ranks the most relevant snippets, resolves conflicts, and answers using only the sources you are permitted to see."
Pages take priority over inferred context. When Genie One answers a question about a concept defined in a Page, "it prioritizes the Page's definition over context it infers automatically, and cites the Page so users can confirm the source."
Permissions apply to inferred context: "Snippets are also gated by Unity Catalog permissions, so Genie only uses snippets extracted from assets you have permission to see." To check an answer, "click the citation icons in a response" to see which knowledge sources Genie used.

Why combine modeled and inferred context?

The Databricks blog post argues for a split: "Humans should explicitly define and govern the small set of concepts that cannot be wrong, things like revenue, compliance rules and core KPIs." The other half is learned: "The broader ontology should continuously learn the long tail from how the organization operates."

How is a context layer different from a semantic layer?

In Databricks' terms, the governed definitions are the modeled half and the context layer adds the inferred half. Unity Catalog semantics is "a collection of tools to help you define standard business metrics, terms, and organizational structures on top of your catalog data". Its components make up "the human-modeled layer of the Genie Ontology: the business context that you define and govern explicitly". Genie One then "combines this modeled context with context it infers automatically from your assets and usage, forming a unified context layer".
Metric views are the core of the modeled layer. They "provide a centralized way to define and manage business metrics by separating measure definitions from the fields (also called dimensions) used to group, filter, and aggregate them".

FAQs

Is Genie Ontology generally available?

No. The docs say "This feature is in Public Preview." Pages, which hold governed business definitions, are in Beta.

Do I have to define everything by hand?

No. Genie "automatically extracts and maintains" inferred context from your existing assets and usage. You define, govern and certify the modeled context: metric views, domains and Pages.

Who can see an inferred snippet?

Only people with permission to see the asset it came from, because snippets are gated by Unity Catalog permissions.

Which Genie experiences use the ontology?

Genie One and Genie Code. The docs say "context you curate once applies to both surfaces."