NVIDIA Auto Ontology is an open-source Python framework for adding a semantic context layer to enterprise data so AI agents can answer questions more reliably. Its premise is that access to documents and databases is not enough: agents also need consistent definitions of business terms, knowledge of how entities relate, guidance about which data to use, and evidence showing where answers came from. The framework builds and connects ontologies from enterprise metadata, reconciles conflicting definitions across systems, and supports learning from certified corrections and usage. It aims to improve retrieval and make analytics more explainable, rather than replace organizations’ existing semantic models.
The software translates natural-language questions into SQL, checks queries against defined business meanings and lineage, and can return provenance with answers. Context is available to agents through MCP, a command-line interface, and an API. Its system includes a web interface, FastAPI services, and a PostgreSQL catalog with pgvector; it uses NVIDIA NIM models for reasoning, entity extraction, embeddings, and reranking. Deployment options include Docker Compose and Kubernetes, with support for Databricks, PostgreSQL, Snowflake, and DuckDB sources. Authentication and role-based access are supported, and Vault can protect database credentials; without Vault, credentials added through the interface are stored in plaintext in PostgreSQL. The project uses the Apache 2.0 license but is not currently accepting external contributions.
Summary generated by AI from the linked article. hn.today is not affiliated with Hacker News or Y Combinator.
Anthropic’s IPO filing says future models may soon meet its definition of recursive self-improvement, while none of its current models have reached that point. Its published data shows Claude leads some research tasks but remains fully supervised, with humans still choosing research goals and evaluating results. (epyhia.com)
Astra improved Qwen’s math performance by running the network for 50 passes and editing its weights, not by simply inspecting raw weights. Similar experiments used different methods, and an attempt to improve Gemma 3 failed. (honnibal.dev)
Itch.io is expanding its filters for games and other content disclosed as using generative AI. The filters will appear in more places across the site. (bsky.app)
Sam Altman argues that people should use less RAM so memory can be redirected to AI data centers. Critics say OpenAI’s efforts to secure a large share of global RAM supplies could also drive up prices and question whether efficiency justifies restricting personal use. (techdirt.com)
Harness engineering improves coding-agent reliability by building better systems for documentation, context loading, validation, and feedback instead of relying on better prompts. The excerpt also outlines an AI testing strategy for agent systems, retrieval-augmented systems, and external integrations. (deepsense.ai)
LLMs often give outdated or incorrect product prices, even when they use web search. The tutorial shows how to build a Python agent that grounds answers in live data and lets code insert verified prices. (blog.apify.com)
Today's best Hacker News stories, summarized and screenshotted, one email a day.