BuildTube Preview About
Back to the feed

Auto-Ontology: Give AI agents checked business meanings for your database

Added

Made by NVIDIA Corporation

View profile on GitHub (opens in a new tab)

Help AI agents answer questions about company databases using reviewed definitions and SQL, with each answer showing its steps.

For developers building data agents.

What it does

Auto-Ontology is NVIDIA's open context layer, a service you run next to your databases to give AI agents context. It compiles database metadata into an ontology: business terms mapped to columns, reviewed SQL for measures, and the relationships between tables. Agents query through it while people review it in a web page.

How to use it

Follow the Docker Compose quickstart: clone the repository, fill in the settings file, load the sample database, start the stack, open the web app, turn on Semantic Compilation in Settings, then ask a question in Chat.

Common use cases

  • Ask in Chat for the revenue of one product in one state and month, and you get the answer together with its SQL and the steps used.
  • Ask about advertising spend in a database that has none, and the agent checks answerability and asks you to clarify instead of guessing.
  • Generate an ontology from your database metadata, and you see business terms you can review, edit and certify in the web interface.

Can I use this?

Runs in Docker on your computer · needs an AI model key

Requirements It runs on your own computer through Docker Compose, with a Postgres sample database loaded by a seed script from a companion repository. Licensed under Apache-2.0.

Caveat If the web page loads without its left menu after a restart, the browser is holding an old session cookie; clear the site data for localhost and sign in again.

Worth knowing The quickstart walks through one sample sales database, Wide World Importers, and Compose publishes Postgres, pgAdmin and the web app on the host, so default passwords must be replaced.

Why it's here

  • 9,044 people starred it on GitHub this week — 10,003 in total.

Numbers as of 8 November 2026, from BuildTube’s latest daily refresh.

Behind it

View source

BuildTube has not run or verified this project. Everything above is written from what the creator published.

Made this?

Quick questions

What is Auto-Ontology?
Auto-Ontology is NVIDIA's open context layer, a service you run next to your databases to give AI agents context.
What can I use Auto-Ontology for?
Ask in Chat for the revenue of one product in one state and month, and you get the answer together with its SQL and the steps used. Ask about advertising spend in a database that has none, and the agent checks answerability and asks you to clarify instead of guessing. Generate an ontology from your database metadata, and you see business terms you can review, edit and certify in the web interface.
Can I run Auto-Ontology on my computer?
It runs on your own computer or server after a local setup.
What do I need to use Auto-Ontology?
Docker with Compose v2, uv, and an NVIDIA API key from build.nvidia.com. The quickstart also has you set new passwords for Postgres, pgAdmin and the admin login.
What license does Auto-Ontology use?
Licensed under Apache-2.0.

More from this week

Related projects

Explore by topic

Post this

A ready-made picture of this card for X, LinkedIn or a group chat. Link to this page alongside it.

Share image for Auto-OntologyDownload share image

For makers

Made this? Show it’s on BuildTube — add the badge to your README or embed the card on your site.

Featured on BuildTube

README badge (Markdown)

[![Featured on BuildTube](https://buildtube.app/badge/gh-nvidia-auto-ontology.svg)](https://buildtube.app/p/gh-nvidia-auto-ontology/)

Embed card (HTML)

<iframe src="https://buildtube.app/embed/gh-nvidia-auto-ontology/" width="560" height="240" style="border:0;max-width:100%" loading="lazy" title="Auto-Ontology on BuildTube"></iframe>

Back to the feed