NVIDIA open-sources Auto Ontology, a semantic layer that makes LLMs understand enterprise data
aigclink · x · 2026-11-04
NVIDIA released Auto Ontology, an open-source Python framework that builds semantic, graph-based reasoning layers over enterprise data so LLMs and agents can answer business queries with certified, auditable definitions.
Problems it solves
- Models can query tables but don't grasp business semantics: asking "last quarter's East China revenue" requires knowing how revenue is computed, which regions count as East China, which date field to use, and which tables can be joined
- Analysts waste time repeatedly explaining metric definitions; these tacit rules can be solidified into a queryable, auditable semantic layer
Core workflow
- Discover: read DB metadata to auto-detect entities, fields, relations
- Ground: map business terms (customer, order, revenue) to concrete tables/columns
- Certify: business users review and certify terms, SQL metrics, table relations
- Validate: pre-query checks whether data actually supports the question (e.g. requests clarification instead of computing ROI without ad-spend data)
The repo has drawn 3.1k stars; it turns analysts' implicit knowledge into AI-callable infrastructure so sales/finance/ops agents share one certified business vocabulary.
More from coding & agent
- Dev proposes LLM-generated ultra-rare content to build anime-style MMORPGs — rickasaurus · 2026-11-04
- Dev builds Chrome extension to bring back the classic Twitter bird on X — joonasvirtanen · 2026-11-04
- Developer lets his app record 'wishes' and auto-spawn OpenCode background agents to fulfill them — SubtleGradient · 2026-11-04
- Post-train your model on your own agent harness with OpenHands, Harbor, SkyRL and Snowflake — StasBekman · 2026-11-04
- Google Antigravity redesign leaked with context usage and quota indicators — clmt · 2026-11-04
- Microsoft's 4B Mage-VL-Cowork Hits 66% on OSWorld-Verified, Near Human Level — microsoft · 2026-11-04