auto-ontology
Auto Ontology is a Python framework for building semantic, graph-based reasoning layers over data. It enables deterministic query generation, better context retrieval, and more explainable AI for LLM and agent workflows.
- Stars
- 521
- Forks
- 62
- + today
- +28
- Created
- 6mo
Ranking data as of November 3, 2026 (UTC).
Star History
Today, hour by hour
Overview
Auto Ontology is a Python framework for building semantic, graph-based reasoning layers over data. It enables deterministic query generation, better context retrieval, and more explainable AI for LLM and agent workflows. It ranks #6 on GitTiger, gaining +28 stars on November 3, 2026 (UTC).
The project is written in Python and has 62 forks. It was created 6mo ago.
What it is
NVIDIA/auto-ontology is a Python framework, licensed Apache-2.0, that puts a graph-based semantic reasoning layer on top of your data. The repository description pitches it at LLM and agent workflows, promising deterministic query generation, improved context retrieval and answers that are easier to explain.
What the README preview says it does
The preview sorts the project into four verbs: - **Discover** – derive enterprise ontologies from metadata, hook up ontologies and semantic models you already have, and produce knowledge graphs ready for AI use. - **Ground** – supply agents with dependable business meaning for metrics, entities and relationships, and resolve conflicting definitions across systems. - **Validate** – compare generated queries with certified definitions and lineage, attaching provenance to every answer. - **Expose** – make context available to agent harnesses via MCP, a CLI or an API.
The preview ends mid-sentence on natural-language support, so anything beyond that was not supplied.
Getting started
The supplied data holds no install or usage steps. Follow the setup instructions in the repository README at https://github.com/NVIDIA/auto-ontology.
git clone https://github.com/NVIDIA/auto-ontology.git
cd auto-ontology
# see README for setupFrequently Asked Questions
What problem does Auto Ontology target?
According to its description, it adds a semantic, graph-based reasoning layer over data so that LLM and agent workflows get deterministic query generation, better context retrieval and more explainable AI.
How can agents consume it?
The README preview mentions MCP, a CLI and an API as ways to serve context to any agent harness.
What language and licence does it use?
The metadata lists Python as the language and Apache-2.0 as the licence.
How do I install it?
Installation steps were not supplied. Check the repository README.