Skip to content

CONCEPT Cited by 1 source

Data products as AI context

The insight that a data product — not a raw table, not a document, and not schema metadata alone — is the correct unit of context for AI agents reasoning over enterprise data.

Definition

A data product in this framing bundles:

  • Ownership — who is responsible for correctness and SLA
  • Lifecycle state — draft / published / deprecated
  • Domain assignment — which business area the product serves
  • Linked governed assets — tables, views, dashboards under catalog governance
  • Context entries — markdown documentation explaining what the product supports, how tables relate, what caveats matter, and how it should be consumed

The distinction from raw metadata: "Agents need more than schema metadata, and they need more than a document. They need to understand which business questions a data product supports, how the important tables relate, what caveats matter, and which surrounding assets should be used with it." (Source: sources/2026-07-21-databricks-why-rd-data-belongs-in-the-lakehouse-and-why-agents-need-it)

Why this matters for system design

The claim is architectural, not just organisational: the governed context layer that makes a data product trustworthy for a human engineer is exactly the context layer that makes an agent's answer trustworthy. Building the AI context layer is therefore a data-engineering problem (governance, lineage, documentation) before it is a model-selection problem.

Seen in

Last updated · 590 distilled / 1,788 read