NVIDIA has released Auto Ontology, an open-source semantic layer framework under the Apache 2.0 license, aimed at helping AI agents generate SQL queries from certified business definitions rather than raw, potentially inconsistent data. The framework builds a contextual layer from enterprise metadata so agents can work with trusted metrics, relationships, and lineage-aware definitions.
According to the project description cited by Techub, the repository frames the problem plainly: getting a revenue answer quickly has little value if the number is wrong. Auto Ontology is designed around that issue, with five core capabilities: generating ontologies from metadata, supplying agents with trusted metric and relationship definitions, validating generated queries against certified definitions and lineage, serving the system through MCP, CLI, and API interfaces, and feeding certified corrections back into the ontology.
Its stack includes a Next.js frontend, a FastAPI backend, a worker for embeddings, and a Postgres database using pgvector. NVIDIA said testing showed the framework cut ontology creation time by 90%. Auto Ontology is compatible with Databricks, Postgres, Snowflake, and DuckDB, and model endpoints can be configured through environment variables.
NVIDIA has released Auto Ontology, an open-source semantic layer framework under the Apache 2.0 license, according to Techub. The framework builds a contextual layer from enterprise metadata so AI agents can write SQL queries based on certified business definitions.
The project repository states the core issue directly: if the numbers are wrong, getting a revenue answer quickly has little value.
Five core functions
Auto Ontology includes five core capabilities:
- Generating ontologies from metadata;
- Providing agents with trusted metric and relationship definitions;
- Validating generated queries against certified definitions and lineage;
- Serving the framework through MCP, CLI, and API interfaces;
- Feeding certified corrections back into the ontology.
Stack and compatibility
The framework uses a Next.js frontend, a FastAPI backend, a worker for handling embeddings, and a Postgres database with pgvector.
On data source support, Auto Ontology is compatible with Databricks, Postgres, Snowflake, and DuckDB. Model endpoints can also be configured through environment variables.
Testing result
NVIDIA said testing showed the framework reduced ontology creation time by 90%.
This article was originally published by Bit.Fan. For more cryptocurrency news and market insights, visit www.bit.fan. Disclaimer:
The market information, project data, and third-party content displayed on this platform are for industry information sharing only and do not constitute any form of investment advice or return commitment.
Cryptocurrency trading carries high risks. Users should fully assess their risk tolerance and make independent decisions. All profits, losses, and legal responsibilities are borne by the users themselves.
