Skip to main content

Automatic AI Documentation for Data Entities

Updated over a week ago

Overview

Maintaining comprehensive and up-to-date documentation across your entire data stack can be challenging and time-consuming. Foundational addresses this with its Automatic AI Documentation feature, ensuring every data entity—from warehouse tables and BI dashboards to data lake files and upstream operational tables —has meaningful, current documentation with minimal manual effort. This feature helps improve data discovery, understanding, and trust throughout your organization.

How it Works

Foundational employs a multi-step process to document your data assets automatically:

  1. Extracting Existing Documentation: Foundational begins by intelligently searching for and extracting existing documentation from all your connected systems. This includes descriptions native to your data warehouse (e.g., table comments), metadata from BI tools (like dashboard or report descriptions), documentation defined in transformation layers (such as dbt model descriptions in schema.yml files), and even comments associated with data models in the source code of upstream operational systems (e.g., from ORM definitions).

  2. AI Documentation Generation (When Needed): If Foundational cannot find any existing documentation for an entity in any connected source, it automatically generates a detailed description using advanced AI models. This AI leverages the rich context available from Foundational, including the entity's end-to-end lineage (upstream sources, downstream consumers), query string, data types and structure, column names, and even the documentation of related or similar entities, to produce relevant and informative descriptions.
    Every documentation that has been automatically generated by Foundational's AI is clearly marked in the user interface.

  3. Keeping Documentation Updated: Documentation isn't useful if it's stale. Foundational continuously monitors your data stack for changes. Whenever an entity's definition, code, lineage, usage patterns, or other relevant characteristics change, Foundational automatically reviews and updates the any automatically generated documentation if necessary. This ensures your automatic AI documentation remains accurate and trustworthy over time without manual intervention.

Editing Documentation

Users can always review and edit any documentation, including AI-generated descriptions. This allows you to add specific business context, refine the wording, correct inaccuracies, or tailor the description to specific internal standards. Edited descriptions are preserved and prioritized, ensuring your custom input is respected.

Configuration

Automatic AI Documentation is enabled by default for all Foundational customers. There are no configuration steps required on your part to activate this feature. Foundational automatically begins discovering and documenting your connected data assets as soon as integrations are set up.

Did this answer your question?