ForkFree

Data and analytics

auto-ontology

by NVIDIA

Auto Ontology builds a layer of business meaning over database metadata, so AI agents know what metrics and tables mean and where they came from. It lets people ask questions in plain language, turns them into checked SQL, and shows where each answer came from.

Worth a lookEstablished, with a few things to check (listed below).

View on GitHub

Stars
633
Licence
Apache-2.0
Runs on
A small server
Setup
Technical
Language
Python
Age
185 days

What you can do with it

  • Ask questions of a Postgres, Snowflake, Databricks or DuckDB database in plain language and get SQL answers
  • Generate an ontology and catalogue of shared metric and entity definitions from your database metadata
  • Serve that business context to AI agents through MCP, a command line or an API

Make money with it

A freelancer could set up governed natural-language data querying for small companies as a paid consulting service.

An idea, not financial advice. Check the licence before charging for anything built on it.

Ways a business could use it

  • Let staff ask questions of company sales or finance data in plain English without writing SQL, with consistent metric definitions
  • Give an in-house AI assistant a trusted catalogue of data definitions so its reported figures are checked and traceable

Can you use it commercially?

Free to use, change and sell, including in paid products. Keep the copyright notice.

Not legal advice. Read the project's licence file before building a product on it.

Getting it running

Clone the repository, copy .env.example to .env with your keys and database details, then run docker compose up -d --build (a Helm chart is available for Kubernetes).

Follow the official instructions in the project README.

Things to check

  • Needs an NVIDIA API key for the hosted models, or self-hosted NIM models, which may cost money
  • Several services to run (Next.js frontend, two Python services, Postgres), so it is heavy for small setups
  • Project is not accepting external contributions
  • Aimed at enterprise data teams, so it may be more than a small business needs

Who's talking about it

Checked 30 Oct 2026. Summary drafted by AI from the project's README.

More in Data and analytics

Scrapling D4Vinci
Worth a look

Scrapling is a Python library for collecting data from websites, covering single page requests, JavaScript-heavy sites and large crawls. It is used to pull out text, prices, listings and other information and save it as JSON, CSV or Markdown.

90.4k stars +126 a day BSD-3-Clause
metabase metabase
Check the licence

Metabase is a business intelligence tool that connects to your databases so people can ask questions, build dashboards and share reports without needing to write SQL. It can be self-hosted or used as a paid cloud service, and it can also be embedded in other apps.

Could replace Tableau, Microsoft Power BI

50k stars +36 a day Unclear
daily_stock_analysis ZhuLinsen
Worth a look

A Python system that uses large language models to analyse a personal watchlist of shares across several markets, including China, Hong Kong, the US, Japan, Korea and Taiwan. It gathers market data and news, then sends a daily "decision dashboard" report to chat apps or email.

67.3k stars +44 a day MIT