{"slug": "nvidia-agent-toolkit-expands-grounding-enterprise-agents-in-company-data", "title": "NVIDIA Agent Toolkit Expands, Grounding Enterprise Agents in Company Data", "summary": "NVIDIA expanded its open source NVIDIA Agent Toolkit with the new NVIDIA Agent Knowledge Foundation, built on CUDA-X libraries (cuDF, cuGraph, cuML, cuOpt, cuVS) plus the new Auto-Ontology library, Nemotron Retriever and the Kumo open model family (Kumo Tabular, Kumo Relational, Kumo Time Series, Kumo Forecast), alongside the OpenShell secure agent runtime and NemoClaw blueprints. NVIDIA said Palantir is using the foundation through Foundry, Ontology and AIP for its supply chain organization, Databricks pairs Kumo Anomaly with Genie Ontology, Snowflake offers Kumo Engine and Kumo Relational in Snowflake Notebooks, and Amazon OpenSearch Service improved search relevance with Nemotron embeddings at virtually no added latency. The toolkit is aimed at letting enterprises ground agents in company data without moving data, rebuilding workflows or changing governance controls.", "body_md": "Agents are only as useful as what they know, and only as trustworthy as the boundaries they run within.\n\nLeading enterprises are using the open source NVIDIA Agent Toolkit — now expanded with new tools — to customize and control agents, use company data to inform their answers and actions, monitor what they do and govern how they operate.\n\nNVIDIA Agent Toolkit comprises:\n\n- The new [NVIDIA Agent Knowledge Foundation](https://developer.nvidia.com/topics/ai/agent-knowledge-foundation) with NVIDIA CUDA-X libraries\n- NVIDIA open models including Nemotron Retriever and the new [Kumo](https://blogs.nvidia.com/blog/open-models-europe-industry) family\n- [NVIDIA OpenShell](https://nvidianews.nvidia.com/news/openshell-reference-architecture) secure agent runtime\n- [NVIDIA NemoClaw](https://www.nvidia.com/en-us/ai/nemoclaw/) secure agent blueprints\n- Tools and skills, like NeMo Relay and NeMo Compass\n\nTo develop agents that accomplish real work and are trustworthy from development to deployment, builders need an open agent stack that brings together the models, harness, runtime, data and tools.\n\nNVIDIA Agent Knowledge Foundation helps enterprises turn data into company intelligence agents can use. Agents can find relevant information, understand what it means for the business and use it to make predictions and help people decide what to do next. Companies can use Agent Knowledge Foundation without moving data, rebuilding existing workflows or changing governance controls.\n\nThe foundation features [NVIDIA CUDA-X libraries](https://www.nvidia.com/en-us/technologies/cuda-x/), including cuDF, cuGraph, cuML, cuOpt and cuVS, and the new [Auto-Ontology](https://github.com/NVIDIA/auto-ontology) library, which translates raw company data into familiar business concepts — like customers, products and orders — and maps how they relate. \n\nIt also includes [NVIDIA Nemotron Retriever](https://developer.nvidia.com/topics/ai/nemotron) and [Kumo Structured](https://developer.nvidia.com/topic/ai/agent-knowledge-foundation/kumo/) open models — Kumo Tabular, Kumo Relational, Kumo Time Series and Kumo Forecast — which help agents find relevant company information to answer questions and make predictions. \n\nOpenShell provides the secure runtime boundary, giving enterprises control over what agents can access and do. NVIDIA NemoClaw blueprints provide reference designs for building secure, specialized agents. The newest [NemoClaw](https://build.nvidia.com/nvidia/nemoclaw-for-agent-optimization) blueprint for agent optimization provides a step-by-step starting point for improving agent intelligence — making them better at their jobs over time.\n\n## **Enterprises Adopt NVIDIA Agent Knowledge Foundation**\n\nThe new NVIDIA Agent Knowledge Foundation is already at work across enterprises.\n\nWith Palantir, [NVIDIA’s supply chain organization](https://nvidianews.nvidia.com/news/nvidia-and-palantir-bring-sovereign-intelligence-to-critical-supply-chains) is using Agent Knowledge Foundation through Palantir Foundry, Ontology and AIP to deploy specialized agents, create actionable intelligence from proprietary data, predict outcomes and optimize decision-making.\n\n[Databricks](http://databricks.com/blog/databricks-ai-runtime-nvidia-train-fine-tune-models) pairs NVIDIA Kumo Anomaly with Databricks Genie Ontology to power context-aware agents.\n\n[Snowflake](https://www.snowflake.com/en/blog/nvidia-kumo-graph-ml-snowflake/) is enabling customers to use the NVIDIA Kumo Engine and Kumo Relational open models in their Snowflake Notebook environments to unlock more predictive capabilities.\n\nNVIDIA Nemotron Retriever models make search faster and more accurate. Using the embedding model, [Amazon OpenSearch Service](https://opensearch.org/blog/nvidia-nemotron-embeddings-for-agentic-search-in-opensearch/) improved relevance with fewer search steps and virtually no added latency. [SAP](https://community.sap.com/t5/technology-blog-posts-by-sap/sap-and-nvidia-training-compact-instruction-following-rerankers-using-opd/ba-p/14506708)’s fine-tuned reranker ran nearly 3x faster. ServiceNow improved reranking quality while cutting latency by nearly a quarter. Zoom is using Nemotron Retriever models to improve document understanding, search relevance and context retrieval for Zoom AI Services. \n\nAmazon Web Services (AWS), ClickHouse, Cloudera, [Cohesity](https://www.cohesity.com/blogs/bring-relevant-enterprise-context-to-ai-agents-with-cohesity-and-nvidia/), [Dassault Systèmes](https://blog.3ds.com/topics/company-news/long-running-agents), Dell, [Elastic](https://www.elastic.co/blog/elastic-nvidia-cuvs-integration), Exa, IBM, NetApp, Oracle, Perplexity, [Starburst](http://starburst.io/press-release/starburst-introduces-torin-gpu-native-analytics-engine-built-on-nvidia-cudf) and [VAST Data](http://www.vastdata.com/blog/building-semantic-enterprise-search-for-ai-agents) are also among those building on Agent Knowledge Foundation. \n\n*Learn more by joining this* *GTC Berlin session on building, deploying and optimizing specialized agents in production**.*", "url": "https://wpnews.pro/news/nvidia-agent-toolkit-expands-grounding-enterprise-agents-in-company-data", "canonical_source": "https://blogs.nvidia.com/blog/enterprise-platforms-agent-toolkit/", "published_at": "2026-10-21 10:10:01+00:00", "updated_at": "2026-10-21 11:30:32.045674+00:00", "lang": "en", "topics": ["ai-agents", "ai-tools", "large-language-models", "ai-infrastructure", "developer-tools"], "entities": ["NVIDIA", "NVIDIA Agent Toolkit", "NVIDIA Agent Knowledge Foundation", "Nemotron Retriever", "Kumo", "NVIDIA OpenShell", "NVIDIA NemoClaw", "Palantir"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/nvidia-agent-toolkit-expands-grounding-enterprise-agents-in-company-data", "markdown": "https://wpnews.pro/news/nvidia-agent-toolkit-expands-grounding-enterprise-agents-in-company-data.md", "text": "https://wpnews.pro/news/nvidia-agent-toolkit-expands-grounding-enterprise-agents-in-company-data.txt", "jsonld": "https://wpnews.pro/news/nvidia-agent-toolkit-expands-grounding-enterprise-agents-in-company-data.jsonld"}}